https://linux-kvm.org/api.php?action=feedcontributions&feedformat=atom&user=Lmr 2026-04-05T23:02:15Z User contributions MediaWiki 1.39.5 https://linux-kvm.org/index.php?title=KVM-Autotest&diff=4583 2012-09-29T20:18:23Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest =<br /> <br /> KVM autotest is the former name of a set of virtualization tests that include KVM and other virtualization backends on Linux platforms. Now the project is known as virt-test, and the current documentation can be seen on the following URL: <br /> <br /> https://github.com/autotest/virt-test/wiki<br /> <br /> If you want to download the suite, go to its main page:<br /> <br /> https://github.com/autotest/virt-test</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=4501 2012-01-03T20:04:02Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list - DEPRECATED =<br /> <br /> KVM autotest is the set of kvm tests implemented on top of the [http://autotest.github.com/ autotest] framework. The current documentation can be seen on the [https://github.com/autotest/autotest/wiki/KVMAutotest autotest wiki], please refer to it. We'll port the remaining items to the autotest issue tracker.<br /> <br /> == Pradeep ==<br /> <br /> * SLES support - <br /> ** Owner - Onkar n Mahajan (currently working)<br /> ** Expected delivery to community - 1-Dec-2011<br /> <br /> * virtio-blk disk serial # test (extend physical_resource_check test)<br /> ** Owner - <br /> ** Expected delivery to community-<br /> <br /> * enhance virtio-blk hotplug test<br /> ** virsh based attach-disk/detach disk<br /> ** enumerate cloud scenarios (live add, reboot after, shutdown, etc.)<br /> ** migration after live add<br /> ** Owner -<br /> ** Expected delivery to community- 1-Oct-2011<br /> <br /> * Add use QED image format during tests.<br /> ** Owner - <br /> ** Expected delivery to community- 1-Dec-2011<br /> <br /> * virsh stream.<br /> ** Owner -<br /> ** Expected delivery to community-<br /> <br /> * pci-multi function testing (&gt;32 virtio devices, disk and network)<br /> ** test out max number of virtio blk and nic devices<br /> ** Owner -<br /> ** Expected delivery to community-<br /> <br /> * virtfs patch inclusion to autotest. <br /> ** Owner -<br /> ** Expected delivery to community-<br /> <br /> * Windows support (ping, file_transfer, ... network tests needs to be added)<br /> ** Owner -<br /> ** Expected delivery to community-<br /> <br /> * Svirt - sVirt is a community effort aimed at integrating Mandatory Access Control (MAC) security and Linux-based virtualization (KVM, lguest et al). <br /> ** Owner - Onkar N Mahajan (currently working)<br /> ** Expected delivery to community- 1-Nov-2011<br /> <br /> * &quot;Migration + other task with parallel &quot;<br /> ** Owner -<br /> ** Expected delivery to community-<br /> <br /> * Network rate limiting<br /> ** Owner -<br /> ** Expected delivery to community-<br /> <br /> * Fix old autoetst patches/bugs<br /> ** Owner -<br /> ** Expected delivery to community-<br /> <br /> * Include libvirt_cim in autotest<br /> ** Owner - <br /> ** Expected delivery to community- <br /> <br /> * Power Support<br /> ** Owner - IBM<br /> ** Expected Delivery to the community - Oct. 2011</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest-Docs&diff=4496 2012-01-03T19:57:57Z

<p>Lmr: </p> <hr /> <div>KVM autotest is the set of kvm tests implemented on top of the [http://autotest.github.com/ autotest] framework. The current documentation can be seen on the [https://github.com/autotest/autotest/wiki/KVMAutotest autotest wiki], please refer to it.</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest&diff=4494 2012-01-03T19:23:47Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest =<br /> <br /> KVM autotest is the set of kvm tests implemented on top of the [http://autotest.github.com/ autotest] framework. The current documentation can be seen on the [https://github.com/autotest/autotest/wiki/KVMAutotest autotest wiki], please refer to it.</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3671 2011-07-26T18:46:56Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> == lmr ==<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> <br /> === done ===<br /> <br /> == akong ==<br /> <br /> * Turn tsc_drift into a kvm unit test.<br /> * push redhat internal kvm subtests to upstream.<br /> The ones left:<br /> ** hdparm.py<br /> ** cpuflags.py<br /> ** qmp_command.py<br /> ** qmp_event_notification.py<br /> ** floppy.py<br /> ** format_disk.py<br /> ** usb.py<br /> <br /> == jason ==<br /> <br /> * Port network tests into windows<br /> * Profiler of pidstat<br /> * guest-&gt;guest guest-&gt;host netperf support<br /> * More readable test result of netperf<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Cleber ==<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3650 2011-06-29T12:33:52Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> == lmr ==<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> <br /> === done ===<br /> <br /> == akong ==<br /> <br /> * push redhat internal kvm subtests to upstream. The ones left:<br /> ** hdparm.py<br /> ** cpuflags.py<br /> ** nmi_watchdog.py<br /> ** qmp_command.py<br /> ** qmp_event_notification.py<br /> ** cdrom.py<br /> ** floppy.py<br /> ** format_disk.py<br /> ** usb.py<br /> <br /> == jason ==<br /> <br /> * Port network tests into windows<br /> * Profiler of pidstat<br /> * guest-&gt;guest guest-&gt;host netperf support<br /> * More readable test result of netperf<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Cleber ==<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3649 2011-06-29T03:31:43Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> == lmr ==<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> <br /> === done ===<br /> <br /> == akong ==<br /> <br /> * push redhat internal kvm subtests to upstream. The ones left:<br /> ** hdparm.py<br /> ** cpuflags.py<br /> ** nmi_watchdog.py<br /> ** qmp_command.py<br /> ** qmp_event_notification.py<br /> ** cdrom.py<br /> ** floppy.py<br /> ** format_disk.py<br /> ** lvm.py<br /> ** usb.py<br /> <br /> == jason ==<br /> <br /> * Port network tests into windows<br /> * Profiler of pidstat<br /> * guest-&gt;guest guest-&gt;host netperf support<br /> * More readable test result of netperf<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Cleber ==<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3648 2011-06-29T03:30:52Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> === Done ===<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> <br /> === done ===<br /> <br /> == akong ==<br /> <br /> * push redhat internal kvm subtests to upstream. The ones left:<br /> ** hdparm.py<br /> ** cpuflags.py<br /> ** nmi_watchdog.py<br /> ** qmp_command.py<br /> ** qmp_event_notification.py<br /> ** cdrom.py<br /> ** floppy.py<br /> ** format_disk.py<br /> ** lvm.py<br /> ** usb.py<br /> <br /> == jason ==<br /> <br /> * Port network tests into windows<br /> * Profiler of pidstat<br /> * guest-&gt;guest guest-&gt;host netperf support<br /> * More readable test result of netperf<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Cleber ==<br /> <br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3647 2011-06-29T03:22:18Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> === Done ===<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> <br /> === done ===<br /> <br /> == akong ==<br /> <br /> * push redhat internal kvm subtests(more than 18) to upstream<br /> ''hdparm.py cpuflags.py multi_disk.py nmi_watchdog.py qmp_command.py qmp_event_notification.py cdrom.py fillup_disk.py floppy.py format_disk.py lvm.py usb.py ... ''[doing..]<br /> stop_continue.py[review..]<br /> image_copy.py module_probe.py[done]<br /> <br /> == jason ==<br /> <br /> * Port network tests into windows<br /> * Profiler of pidstat<br /> * guest-&gt;guest guest-&gt;host netperf support<br /> * More readable test result of netperf<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Cleber ==<br /> <br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest&diff=3547 2011-04-29T02:23:02Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest =<br /> <br /> The KVM-Autotest system is a client test for the [http://autotest.kernel.org/ autotest] framework. Its main purpose is to serve as an automated regression testing tool for KVM developers, and for doing regular automated testing of KVM (provided you use it with the server testing infrastructure).<br /> <br /> Autotest is a project that aims to provide tools and libraries to perform automated testing on the linux platform. KVM autotest is a subproject under the autotest umbrella. For more information on autotest, see http://autotest.kernel.org/. Autotest is a modular framework, and KVM autotest is nothing more than a submodule of the client module. Of course, for more complex testing we also use other parts of autotest, and for centralized tests, we deploy a full autotest infrastructure.<br /> <br /> KVM autotest aims to be a centralizing project for most of the KVM functional and performance testing needs. We cover:<br /> <br /> * Guest OS install, for both Windows (WinXP - Win7) and Linux (RHEL, Fedora, OpenSUSE and others through step engine mechanism)<br /> * Serial output for Linux guests<br /> * Migration, networking, timedrift and other types of tests<br /> * Monitor control for both human and QMP protocols<br /> * Build and use qemu using various methods (source tarball, git repo, rpm)<br /> * Some level of performance testing can be made.<br /> * The KVM unit tests can be run comfortably from inside kvm autotest, we do have full integration with the unittest execution<br /> <br /> We support x86_64 hosts with hardware virtualization support (AMD and Intel), and Intel 32 and 64 bit guest operating systems. <br /> <br /> For an overview about KVM autotest, how this project was created, its goals, structure, and how to develop simple tests, you can refer to the [http://vimeo.com/15234614 KVM forum 2010 presentation], [[Media:2010-forum-Kvm-autotest.pdf | Slides]]<br /> <br /> * [[KVM-Autotest/Getting Started|Getting Started with KVM testing]] - How to get your first tests up and running.<br /> * [[KVM-Autotest/Cmdline|Using the autotest command line]] - How to use the autotest cli to run your kvm tests in an easier way<br /> * [[KVM-Autotest/Test Config File|The test configuration file format]] - How to work with the framework configuration file format. <br /> <br /> * [[KVM-Autotest/Tests|KVM subtest reference]] - Brief overview of the available test API.<br /> * [[KVM-Autotest/Contributing|Contribute to KVM autotest]] - Guidance on how to submit patches. <br /> <br /> * [[KVM-Autotest/Parameters|Framework parameters]] - Lists and explains the parameters used by the framework.<br /> <br /> * [[KVM-Autotest/AddNewGuests|Adding New Step files]] - The default install mode for most guests now is the unattended install, but for some guests, we might need to use step based install. This explains how to do so.<br /> <br /> ----<br /> <br /> * [[KVM-Autotest/TODO|TODO]] - TODO list, for those interested in helping.</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3546 2011-04-29T02:22:31Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Generate autotest API documentation with tools such as doxygen<br /> * Hotplug tests during migration<br /> * Work on setting up the conmux server for our internal test grid<br /> * Research how to trigger jobs based on events rather than using the web interface setting recurrent jobs<br /> * Set up a qemu-block job, with kwolf's qemu-block tree<br /> * Set up a vhost job, with the upstream job (mst's trees for kernel and userspace)<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Done ===<br /> <br /> * Infrastructure for testing different vm storage backends: file/lvm/iSCSI - Done by Beijing team<br /> * Made some internal patches appropriate for upstream<br /> * Copy kvm-autotest@redhat.com on the email list of results<br /> * Send code to retrieve host kernels based on koji/brew tags upstream<br /> * Resolved UUID mount point check problem<br /> * QMP suite integrated upstream<br /> * Wrote code to retrieve kernels based on koji/brew tags<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> <br /> === done ===<br /> <br /> * Post-commit review of new network patches (committed on Oct 7 2010)<br /> * Use more exceptions in utility functions:<br /> ** Add convenience functions to kvm_subprocess.py to make running shell commands shorter<br /> ** Add convenience functions to the VM class<br /> * Refactor _get_command_output() and friends in kvm_monitor.py and add cmd_raw(), cmd_obj() and cmd_qmp() as required by Luiz Capitulino's test suite<br /> * Use select() instead of sleep() in kvm_monitor.py<br /> * Simplify migration code if possible (in kvm_vm.py and kvm_test_utils.py)<br /> * Fix VM.get_ifname() somehow (uses self.vnc_port -- bad)<br /> <br /> == akong ==<br /> <br /> <br /> * push redhat internal kvm subtests(more than 18) to upstream<br /> ''trans_hugepage_defrag.py trans_hugepage.py trans_hugepage_swapping.py hdparm.py cpuflags.py multi_disk.py nmi_watchdog.py qmp_command.py qmp_event_notification.py cdrom.py fillup_disk.py floppy.py format_disk.py lvm.py usb.py ... ''[doing..]<br /> stop_continue.py[review..]<br /> image_copy.py module_probe.py[done]<br /> <br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge [sent v2]<br /> * Add a new test: check if guest transmits packets when link is up/down [done]<br /> * Report problem with multiple NICs and MAC address tracking [problem fixed]<br /> * review and test michael's whql patchsets, give feedback to michael [done]<br /> * Bonding test [done]<br /> <br /> == jason ==<br /> <br /> * Tests in parallel with migration (reboot/unattended installation/file transfer/ autotest client tests / guest_script,autoit? / memtest86+(steps?)), need modifications to framework?<br /> * Refactor the network cmdline generation, using peer when possible and vlan for elder qemu-kvm<br /> * Port network tests into windows<br /> * Profiler of pidstat<br /> * guest-&gt;guest guest-&gt;host netperf support<br /> * More readable test result of netperf<br /> <br /> === done ===<br /> * Migration to exec test suggested by mst<br /> * Clean the Migration tests: mainly for exec and add offline support<br /> * Write a 'cli' wrapper to make autotest handy/easier for developers<br /> * Refactor hotplug test using qdev and netdev_del/drive_del<br /> <br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Cleber ==<br /> <br /> * Create initscripts for the scheduller daemon (monitor_db_babysitter)<br /> * Revisit job parameterization to ease job submission<br /> * Integrate Brew/Koji messaging with automatic job submission<br /> <br /> === done ===<br /> <br /> * Look and fix the block_hotplug test upstream<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest&diff=3545 2011-04-29T02:20:56Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest =<br /> <br /> The KVM-Autotest system is a client test for the [http://autotest.kernel.org/ autotest] framework. Its main purpose is to serve as an automated regression testing tool for KVM developers, and for doing regular automated testing of KVM (provided you use it with the server testing infrastructure).<br /> <br /> Autotest is a project that aims to provide tools and libraries to perform automated testing on the linux platform. KVM autotest is a subproject under the autotest umbrella. For more information on autotest, see http://autotest.kernel.org/. Autotest is a modular framework, and KVM autotest is nothing more than a submodule of the client module. Of course, for more complex testing we also use other parts of autotest, and for centralized tests, we deploy a full autotest infrastructure.<br /> <br /> KVM autotest aims to be a centralizing project for most of the KVM functional and performance testing needs. We cover:<br /> <br /> * Guest OS install, for both Windows (WinXP - Win7) and Linux (RHEL, Fedora, OpenSUSE and others through step engine mechanism)<br /> * Serial output for Linux guests<br /> * Migration, networking, timedrift and other types of tests<br /> * Monitor control for both human and QMP protocols<br /> * Build and use qemu using various methods (source tarball, git repo, rpm)<br /> * Some level of performance testing can be made.<br /> * The KVM unit tests can be run comfortably from inside kvm autotest, we do have full integration with the unittest execution<br /> <br /> We support x86_64 hosts with hardware virtualization support (AMD and intel), and intel 32 and 64 bits operating systems for the guests.<br /> <br /> For an overview about KVM autotest, how this project was created, its goals, structure, and how to develop simple tests, you can refer to the [http://vimeo.com/15234614 KVM forum 2010 presentation], [[Media:2010-forum-Kvm-autotest.pdf | Slides]]<br /> <br /> * [[KVM-Autotest/Getting Started|Getting Started with KVM testing]] - How to get your first tests up and running.<br /> * [[KVM-Autotest/Cmdline|Using the autotest command line]] - How to use the autotest cli to run your kvm tests in an easier way<br /> * [[KVM-Autotest/Test Config File|The test configuration file format]] - How to work with the framework configuration file format. <br /> <br /> * [[KVM-Autotest/Tests|KVM subtest reference]] - Brief overview of the available test API.<br /> * [[KVM-Autotest/Contributing|Contribute to KVM autotest]] - Guidance on how to submit patches. <br /> <br /> * [[KVM-Autotest/Parameters|Framework parameters]] - Lists and explains the parameters used by the framework.<br /> <br /> * [[KVM-Autotest/AddNewGuests|Adding New Step files]] - The default install mode for most guests now is the unattended install, but for some guests, we might need to use step based install. This explains how to do so.<br /> <br /> ----<br /> <br /> * [[KVM-Autotest/TODO|TODO]] - TODO list, for those interested in helping.</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest&diff=3540 2011-04-29T02:01:08Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest =<br /> <br /> The KVM-Autotest system is a client test for the [http://autotest.kernel.org/ autotest] framework. Its main purpose is to serve as an automated regression testing tool for KVM developers, and for doing regular automated testing of KVM (provided you use it with the server testing infrastructure).<br /> <br /> Autotest is a project that aims to provide tools and libraries to perform automated testing on the linux platform. KVM autotest is a subproject under the autotest umbrella. For more information on autotest, see http://autotest.kernel.org/. Autotest is a modular framework, and KVM autotest is nothing more than a submodule of the client module. Of course, for more complex testing we also use other parts of autotest, and for centralized tests, we deploy a full autotest infrastructure.<br /> <br /> KVM autotest aims to be a centralizing project for most of the KVM functional and performance testing needs. We cover:<br /> <br /> * Guest OS install, for both Windows (WinXP - Win7) and Linux (RHEL, Fedora, OpenSUSE and others through step engine mechanism)<br /> * Serial output for Linux guests<br /> * Migration, networking, timedrift and other types of tests<br /> * Monitor control for both human and QMP protocols<br /> * Build and use qemu using various methods (source tarball, git repo, rpm)<br /> * Some level of performance testing can be made.<br /> * The KVM unit tests can be run comfortably from inside kvm autotest, we do have full integration with the unittest execution<br /> <br /> We support x86_64 guests with hardware virtualization support (AMD and intel), and intel 32 and 64 bits operating systems.<br /> <br /> For an overview about KVM autotest, how this project was created, its goals, structure, and how to develop simple tests, you can refer to the [http://vimeo.com/15234614 KVM forum 2010 presentation], [[Media:2010-forum-Kvm-autotest.pdf | Slides]]<br /> <br /> * [[KVM-Autotest/Getting Started|Getting Started with KVM testing]] - How to get your first tests up and running.<br /> * [[KVM-Autotest/Cmdline|Using the autotest command line]] - How to use the autotest cli to run your kvm tests in an easier way<br /> * [[KVM-Autotest/Test Config File|The test configuration file format]] - How to work with the framework configuration file format. <br /> <br /> * [[KVM-Autotest/Tests|KVM subtest reference]] - Brief overview of the available test API.<br /> * [[KVM-Autotest/Contributing|Contribute to KVM autotest]] - Guidance on how to submit patches. <br /> <br /> * [[KVM-Autotest/Parameters|Framework parameters]] - Lists and explains the parameters used by the framework.<br /> <br /> * [[KVM-Autotest/AddNewGuests|Adding New Step files]] - The default install mode for most guests now is the unattended install, but for some guests, we might need to use step based install. This explains how to do so.<br /> <br /> ----<br /> <br /> * [[KVM-Autotest/TODO|TODO]] - TODO list, for those interested in helping.</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest&diff=3535 2011-04-27T12:36:38Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest =<br /> <br /> The KVM-Autotest system is a client test for the [http://autotest.kernel.org/ autotest] framework. Its main purpose is to serve as an automated regression testing tool for KVM developers, and for doing regular automated testing of KVM (provided you use it with the server testing infrastructure).<br /> <br /> * [[KVM-Autotest/Getting Started|Getting Started with KVM testing]] - How to get your first tests up and running.<br /> * [[KVM-Autotest/Cmdline|Using the autotest command line]] - How to use the autotest cli to run your kvm tests in an easier way<br /> * [[KVM-Autotest/Test Config File|The test configuration file format]] - How to work with the framework configuration file format. <br /> <br /> * [[KVM-Autotest/Tests|KVM subtest reference]] - Brief overview of the available test API.<br /> * [[KVM-Autotest/Contributing|Contribute to KVM autotest]] - Guidance on how to submit patches. <br /> <br /> * [[KVM-Autotest/Parameters|Framework parameters]] - Lists and explains the parameters used by the framework.<br /> <br /> * [[KVM-Autotest/AddNewGuests|Adding New Step files]] - The default install mode for most guests now is the unattended install, but for some guests, we might need to use step based install. This explains how to do so.<br /> <br /> ----<br /> <br /> * [[KVM-Autotest/TODO|TODO]] - TODO list, for those interested in helping.</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest&diff=3534 2011-04-27T12:36:00Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest =<br /> <br /> The KVM-Autotest system is a client test for the [http://autotest.kernel.org/ autotest] framework. Its main purpose is to serve as an automated regression testing tool for KVM developers, and for doing regular automated testing of KVM (provided you use it with the server testing infrastructure).<br /> <br /> * [[KVM-Autotest/Getting Started|Getting Started with KVM testing]] - How to get your first tests up and running.<br /> * [[KVM-Autotest/CmdLine|Using the autotest command line]] - How to use the autotest cli to run your kvm tests in an easier way<br /> * [[KVM-Autotest/Test Config File|The test configuration file format]] - How to work with the framework configuration file format. <br /> <br /> * [[KVM-Autotest/Tests|KVM subtest reference]] - Brief overview of the available test API.<br /> * [[KVM-Autotest/Contributing|Contribute to KVM autotest]] - Guidance on how to submit patches. <br /> <br /> * [[KVM-Autotest/Parameters|Framework parameters]] - Lists and explains the parameters used by the framework.<br /> <br /> * [[KVM-Autotest/AddNewGuests|Adding New Step files]] - The default install mode for most guests now is the unattended install, but for some guests, we might need to use step based install. This explains how to do so.<br /> <br /> ----<br /> <br /> * [[KVM-Autotest/TODO|TODO]] - TODO list, for those interested in helping.</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3520 2011-03-16T11:38:51Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> * Autotest refactor for usage with libvirt/xen. See more detauls at the [[KVM Autotest Refactor page]]<br /> <br /> <br /> * Complete James Ren patch to parametrized control files<br /> * First commit found on 4720 - http://autotest.kernel.org/changeset/4720<br /> * The reason is that we should implement things like job submission and the kvm autotest wrapper properly on top of the parametrized control file feature, that is currently halfway implemented<br /> <br /> Tasks:<br /> <br /> ** Move kvm autotest libraries to client/bin directory, so every client side test can use <br /> ** Review xen-autotest to see what are the differences between our vm class and theirs. Hopefully in the future, xen-autotest and libvirt-autotest will be only different vm implementations of our virt-autotest framework :)<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Generate autotest API documentation with tools such as doxygen<br /> * Walk through autotest with Ray Chauvin<br /> <br /> * Multi host migration - Already have a 1st patchset available, working to rebase it against a cleanup patchset made by Jason<br /> <br /> * Create 'maintenance jobs' just to keep the machines up to date, with appropriate software installed.<br /> * Hotplug tests during migration<br /> * Work on setting up the conmux server for our internal test grid<br /> * Research how to trigger jobs based on events rather than using the web interface setting recurrent jobs<br /> * Set up a qemu-block job, with kwolf's qemu-block tree<br /> * Set up a vhost job, with the upstream job (mst's trees for kernel and userspace)<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Done ===<br /> <br /> * Infrastructure for testing different vm storage backends: file/lvm/iSCSI - Done by Beijing team<br /> * Made some internal patches appropriate for upstream<br /> * Copy kvm-autotest@redhat.com on the email list of results<br /> * Send code to retrieve host kernels based on koji/brew tags upstream<br /> * Resolved UUID mount point check problem<br /> * QMP suite integrated upstream<br /> * Wrote code to retrieve kernels based on koji/brew tags<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> <br /> === done ===<br /> <br /> * Post-commit review of new network patches (committed on Oct 7 2010)<br /> * Use more exceptions in utility functions:<br /> ** Add convenience functions to kvm_subprocess.py to make running shell commands shorter<br /> ** Add convenience functions to the VM class<br /> * Refactor _get_command_output() and friends in kvm_monitor.py and add cmd_raw(), cmd_obj() and cmd_qmp() as required by Luiz Capitulino's test suite<br /> * Use select() instead of sleep() in kvm_monitor.py<br /> * Simplify migration code if possible (in kvm_vm.py and kvm_test_utils.py)<br /> * Fix VM.get_ifname() somehow (uses self.vnc_port -- bad)<br /> <br /> == akong ==<br /> <br /> * review and test michael's whql patchsets, give feedback to michael<br /> * push redhat internal kvm subtests(18) to upstream<br /> ''trans_hugepage_defrag.py trans_hugepage.py trans_hugepage_swapping.py hdparm.py cpuflags.py multi_disk.py nmi_watchdog.py qmp_command.py qmp_event_notification.py cdrom.py fillup_disk.py floppy.py format_disk.py lvm.py module_probe.py stop_continue.py usb.py image_copy.py''<br /> <br /> * Add a new test: check if guest transmits packets when link is up/down [done]<br /> * Report problem with multiple NICs and MAC address tracking [problem fixed]<br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge [request changed]<br /> * Bonding test [done]<br /> <br /> == jason ==<br /> <br /> * Tests in parallel with migration (reboot/unattended installation/file transfer/ autotest client tests / guest_script,autoit? / memtest86+(steps?)), need modifications to framework?<br /> * Refactor the network cmdline generation, using peer when possible and vlan for elder qemu-kvm<br /> * Port network tests into windows<br /> * Profiler of pidstat<br /> * guest-&gt;guest guest-&gt;host netperf support<br /> * More readable test result of netperf<br /> <br /> === done ===<br /> * Migration to exec test suggested by mst<br /> * Clean the Migration tests: mainly for exec and add offline support<br /> * Write a 'cli' wrapper to make autotest handy/easier for developers<br /> * Refactor hotplug test using qdev and netdev_del/drive_del<br /> <br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Cleber ==<br /> <br /> * Create initscripts for the scheduller daemon (monitor_db_babysitter)<br /> * Revisit job parameterization to ease job submission<br /> * Integrate Brew/Koji messaging with automatic job submission<br /> <br /> === done ===<br /> <br /> * Look and fix the block_hotplug test upstream<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3460 2010-12-22T14:31:15Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> * Autotest refactor for usage with libvirt/xen. See more detauls at the [[KVM Autotest Refactor page]]<br /> <br /> <br /> * Complete James Ren patch to parametrized control files<br /> * First commit found on 4720 - http://autotest.kernel.org/changeset/4720<br /> * The reason is that we should implement things like job submission and the kvm autotest wrapper properly on top of the parametrized control file feature, that is currently halfway implemented<br /> <br /> Tasks:<br /> <br /> ** Move kvm autotest libraries to client/bin directory, so every client side test can use <br /> ** Review xen-autotest to see what are the differences between our vm class and theirs. Hopefully in the future, xen-autotest and libvirt-autotest will be only different vm implementations of our virt-autotest framework :)<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Multi host migration - Already have a 1st patchset available, working to rebase it against a cleanup patchset made by Jason<br /> <br /> * Create 'maintenance jobs' just to keep the machines up to date, with appropriate software installed.<br /> * Hotplug tests during migration<br /> * Work on setting up the conmux server for our internal test grid<br /> * Research how to trigger jobs based on events rather than using the web interface setting recurrent jobs<br /> * Set up a qemu-block job, with kwolf's qemu-block tree<br /> * Set up a vhost job, with the upstream job (mst's trees for kernel and userspace)<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Done ===<br /> <br /> * Infrastructure for testing different vm storage backends: file/lvm/iSCSI - Done by Beijing team<br /> * Made some internal patches appropriate for upstream<br /> * Copy kvm-autotest@redhat.com on the email list of results<br /> * Send code to retrieve host kernels based on koji/brew tags upstream<br /> * Resolved UUID mount point check problem<br /> * QMP suite integrated upstream<br /> * Wrote code to retrieve kernels based on koji/brew tags<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> <br /> === done ===<br /> <br /> * Post-commit review of new network patches (committed on Oct 7 2010)<br /> * Use more exceptions in utility functions:<br /> ** Add convenience functions to kvm_subprocess.py to make running shell commands shorter<br /> ** Add convenience functions to the VM class<br /> * Refactor _get_command_output() and friends in kvm_monitor.py and add cmd_raw(), cmd_obj() and cmd_qmp() as required by Luiz Capitulino's test suite<br /> * Use select() instead of sleep() in kvm_monitor.py<br /> * Simplify migration code if possible (in kvm_vm.py and kvm_test_utils.py)<br /> * Fix VM.get_ifname() somehow (uses self.vnc_port -- bad)<br /> <br /> == akong ==<br /> <br /> * review and test michael's whql patchsets, give feedback to michael<br /> * push redhat internal kvm subtests(18) to upstream<br /> ''trans_hugepage_defrag.py trans_hugepage.py trans_hugepage_swapping.py hdparm.py cpuflags.py multi_disk.py nmi_watchdog.py qmp_command.py qmp_event_notification.py cdrom.py fillup_disk.py floppy.py format_disk.py lvm.py module_probe.py stop_continue.py usb.py image_copy.py''<br /> <br /> * Add a new test: check if guest transmits packets when link is up/down [done]<br /> * Report problem with multiple NICs and MAC address tracking [problem fixed]<br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge [request changed]<br /> * Bonding test [done]<br /> <br /> == jason ==<br /> <br /> * Write a 'cli' wrapper to make autotest handy/easier for developers<br /> * Tests in parallel with migration (reboot/unattended installation/file transfer/ autotest client tests / guest_script,autoit? / memtest86+(steps?)), need modifications to framework?<br /> * Refactor hotplug test using qdev and netdev_del/drive_del<br /> * Refactor the network cmdline generation, using peer when possible and vlan for elder qemu-kvm<br /> * Port network tests into windows<br /> <br /> === done ===<br /> * Migration to exec test suggested by mst<br /> * Clean the Migration tests: mainly for exec and add offline support<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Cleber ==<br /> <br /> * Look and fix the block_hotplug test upstream<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3446 2010-12-15T11:27:23Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> * Autotest refactor for usage with libvirt/xen. See more detauls at the [[KVM Autotest Refactor page]]<br /> <br /> <br /> * Complete James Ren patch to parametrized control files<br /> * First commit found on 4720 - http://autotest.kernel.org/changeset/4720<br /> * The reason is that we should implement things like job submission and the kvm autotest wrapper properly on top of the parametrized control file feature, that is currently halfway implemented<br /> <br /> Tasks:<br /> <br /> ** Move kvm autotest libraries to client/bin directory, so every client side test can use <br /> ** Review xen-autotest to see what are the differences between our vm class and theirs. Hopefully in the future, xen-autotest and libvirt-autotest will be only different vm implementations of our virt-autotest framework :)<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Multi host migration - Already have a 1st patchset available, working to rebase it against a cleanup patchset made by Jason<br /> <br /> * Create 'maintenance jobs' just to keep the machines up to date, with appropriate software installed.<br /> * Hotplug tests during migration<br /> * Work on setting up the conmux server for our internal test grid<br /> * Research how to trigger jobs based on events rather than using the web interface setting recurrent jobs<br /> * Set up a qemu-block job, with kwolf's qemu-block tree<br /> * Set up a vhost job, with the upstream job (mst's trees for kernel and userspace)<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Done ===<br /> <br /> * Infrastructure for testing different vm storage backends: file/lvm/iSCSI - Done by Beijing team<br /> * Made some internal patches appropriate for upstream<br /> * Copy kvm-autotest@redhat.com on the email list of results<br /> * Send code to retrieve host kernels based on koji/brew tags upstream<br /> * Resolved UUID mount point check problem<br /> * QMP suite integrated upstream<br /> * Wrote code to retrieve kernels based on koji/brew tags<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> <br /> === done ===<br /> <br /> * Post-commit review of new network patches (committed on Oct 7 2010)<br /> * Use more exceptions in utility functions:<br /> ** Add convenience functions to kvm_subprocess.py to make running shell commands shorter<br /> ** Add convenience functions to the VM class<br /> * Refactor _get_command_output() and friends in kvm_monitor.py and add cmd_raw(), cmd_obj() and cmd_qmp() as required by Luiz Capitulino's test suite<br /> * Use select() instead of sleep() in kvm_monitor.py<br /> * Simplify migration code if possible (in kvm_vm.py and kvm_test_utils.py)<br /> * Fix VM.get_ifname() somehow (uses self.vnc_port -- bad)<br /> <br /> == akong ==<br /> <br /> * push redhat internal kvm subtests(18) to upstream<br /> ''trans_hugepage_defrag.py trans_hugepage.py trans_hugepage_swapping.py hdparm.py cpuflags.py multi_disk.py nmi_watchdog.py qmp_command.py qmp_event_notification.py cdrom.py fillup_disk.py floppy.py format_disk.py lvm.py module_probe.py stop_continue.py usb.py image_copy.py''<br /> <br /> * Add a new test: check if guest transmits packets when link is up/down<br /> * Report problem with multiple NICs and MAC address tracking [problem fixed]<br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge [request changed]<br /> * Bonding test [done]<br /> <br /> == jason ==<br /> <br /> * Write a 'cli' wrapper to make autotest handy/easier for developers<br /> * Tests in parallel with migration (reboot/unattended installation/file transfer/ autotest client tests / guest_script,autoit? / memtest86+(steps?)), need modifications to framework?<br /> * Refactor hotplug test using qdev and netdev_del/drive_del<br /> * Refactor the network cmdline generation, using peer when possible and vlan for elder qemu-kvm<br /> * Port network tests into windows<br /> <br /> === done ===<br /> * Migration to exec test suggested by mst<br /> * Clean the Migration tests: mainly for exec and add offline support<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3445 2010-12-15T11:26:58Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> * Autotest refactor for usage with libvirt/xen. See more detauls at the [[KVM Autotest Refactor page]]<br /> <br /> <br /> * Complete James Ren patch to parametrized control files<br /> * First commit found on 4720 - http://autotest.kernel.org/changeset/4720<br /> * The reason is that we should implement things like job submission and<br /> the kvm autotest wrapper properly on top of the parametrized control<br /> file feature, that is currently halfway implemented<br /> <br /> Tasks:<br /> <br /> ** Move kvm autotest libraries to client/bin directory, so every client side test can use <br /> ** Review xen-autotest to see what are the differences between our vm class and theirs. Hopefully in the future, xen-autotest and libvirt-autotest will be only different vm implementations of our virt-autotest framework :)<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Multi host migration - Already have a 1st patchset available, working to rebase it against a cleanup patchset made by Jason<br /> <br /> * Create 'maintenance jobs' just to keep the machines up to date, with appropriate software installed.<br /> * Hotplug tests during migration<br /> * Work on setting up the conmux server for our internal test grid<br /> * Research how to trigger jobs based on events rather than using the web interface setting recurrent jobs<br /> * Set up a qemu-block job, with kwolf's qemu-block tree<br /> * Set up a vhost job, with the upstream job (mst's trees for kernel and userspace)<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Done ===<br /> <br /> * Infrastructure for testing different vm storage backends: file/lvm/iSCSI - Done by Beijing team<br /> * Made some internal patches appropriate for upstream<br /> * Copy kvm-autotest@redhat.com on the email list of results<br /> * Send code to retrieve host kernels based on koji/brew tags upstream<br /> * Resolved UUID mount point check problem<br /> * QMP suite integrated upstream<br /> * Wrote code to retrieve kernels based on koji/brew tags<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> <br /> === done ===<br /> <br /> * Post-commit review of new network patches (committed on Oct 7 2010)<br /> * Use more exceptions in utility functions:<br /> ** Add convenience functions to kvm_subprocess.py to make running shell commands shorter<br /> ** Add convenience functions to the VM class<br /> * Refactor _get_command_output() and friends in kvm_monitor.py and add cmd_raw(), cmd_obj() and cmd_qmp() as required by Luiz Capitulino's test suite<br /> * Use select() instead of sleep() in kvm_monitor.py<br /> * Simplify migration code if possible (in kvm_vm.py and kvm_test_utils.py)<br /> * Fix VM.get_ifname() somehow (uses self.vnc_port -- bad)<br /> <br /> == akong ==<br /> <br /> * push redhat internal kvm subtests(18) to upstream<br /> ''trans_hugepage_defrag.py trans_hugepage.py trans_hugepage_swapping.py hdparm.py cpuflags.py multi_disk.py nmi_watchdog.py qmp_command.py qmp_event_notification.py cdrom.py fillup_disk.py floppy.py format_disk.py lvm.py module_probe.py stop_continue.py usb.py image_copy.py''<br /> <br /> * Add a new test: check if guest transmits packets when link is up/down<br /> * Report problem with multiple NICs and MAC address tracking [problem fixed]<br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge [request changed]<br /> * Bonding test [done]<br /> <br /> == jason ==<br /> <br /> * Write a 'cli' wrapper to make autotest handy/easier for developers<br /> * Tests in parallel with migration (reboot/unattended installation/file transfer/ autotest client tests / guest_script,autoit? / memtest86+(steps?)), need modifications to framework?<br /> * Refactor hotplug test using qdev and netdev_del/drive_del<br /> * Refactor the network cmdline generation, using peer when possible and vlan for elder qemu-kvm<br /> * Port network tests into windows<br /> <br /> === done ===<br /> * Migration to exec test suggested by mst<br /> * Clean the Migration tests: mainly for exec and add offline support<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3437 2010-12-08T14:58:22Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> * Autotest refactor for usage with libvirt/xen. See more detauls at the [[KVM Autotest Refactor page]]<br /> <br /> Tasks:<br /> <br /> ** Move kvm autotest libraries to client/bin directory, so every client side test can use <br /> ** Review xen-autotest to see what are the differences between our vm class and theirs. Hopefully in the future, xen-autotest and libvirt-autotest will be only different vm implementations of our virt-autotest framework :)<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Multi host migration - Already have a 1st patchset available, working to rebase it against a cleanup patchset made by Jason<br /> <br /> * Create 'maintenance jobs' just to keep the machines up to date, with appropriate software installed.<br /> * Hotplug tests during migration<br /> * Work on setting up the conmux server for our internal test grid<br /> * Research how to trigger jobs based on events rather than using the web interface setting recurrent jobs<br /> * Set up a qemu-block job, with kwolf's qemu-block tree<br /> * Set up a vhost job, with the upstream job (mst's trees for kernel and userspace)<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Done ===<br /> <br /> * Infrastructure for testing different vm storage backends: file/lvm/iSCSI - Done by Beijing team<br /> * Made some internal patches appropriate for upstream<br /> * Copy kvm-autotest@redhat.com on the email list of results<br /> * Send code to retrieve host kernels based on koji/brew tags upstream<br /> * Resolved UUID mount point check problem<br /> * QMP suite integrated upstream<br /> * Wrote code to retrieve kernels based on koji/brew tags<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> <br /> === done ===<br /> <br /> * Post-commit review of new network patches (committed on Oct 7 2010)<br /> * Use more exceptions in utility functions:<br /> ** Add convenience functions to kvm_subprocess.py to make running shell commands shorter<br /> ** Add convenience functions to the VM class<br /> * Refactor _get_command_output() and friends in kvm_monitor.py and add cmd_raw(), cmd_obj() and cmd_qmp() as required by Luiz Capitulino's test suite<br /> * Use select() instead of sleep() in kvm_monitor.py<br /> * Simplify migration code if possible (in kvm_vm.py and kvm_test_utils.py)<br /> * Fix VM.get_ifname() somehow (uses self.vnc_port -- bad)<br /> <br /> == akong ==<br /> <br /> * push redhat internal kvm subtests(18) to upstream<br /> ''trans_hugepage_defrag.py trans_hugepage.py trans_hugepage_swapping.py hdparm.py cpuflags.py multi_disk.py nmi_watchdog.py qmp_command.py qmp_event_notification.py cdrom.py fillup_disk.py floppy.py format_disk.py lvm.py module_probe.py stop_continue.py usb.py image_copy.py''<br /> <br /> * Add a new test: check if guest transmits packets when link is up/down<br /> * Report problem with multiple NICs and MAC address tracking [problem fixed]<br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge [request changed]<br /> * Bonding test [done]<br /> <br /> == jason ==<br /> <br /> * Write a 'cli' wrapper to make autotest handy/easier for developers<br /> * Tests in parallel with migration (reboot/unattended installation/file transfer/ autotest client tests / guest_script,autoit? / memtest86+(steps?)), need modifications to framework?<br /> * Clean the Migration tests: mainly for exec and add offline support<br /> * Refactor hotplug test using qdev and netdev_del/drive_del<br /> * Refactor the network cmdline generation, using peer when possible and vlan for elder qemu-kvm<br /> * Port network tests into windows<br /> * Migration to exec test suggested by mst<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM_Autotest_Refactor_page&diff=3433 2010-12-08T01:16:02Z

<p>Lmr: </p> <hr /> <div>= KVM autotest refactor =<br /> <br /> == Suggested approach ==<br /> <br /> 1) Move code that has negligible impact [see table below] from 'client/tests/kvm' to 'client' ASAP<br /> <br /> + This would immediatelly benefit xen autotest, since it is frequently lagging behing features and fixes. One example is newer syntax in kvm_config.py that was not available in the xen autotest counterpart.<br /> <br /> 2) Synchronize that code (step 1) usage between kvm and xen autotest<br /> + Rewrite current tests, changing imports of modules from 'kvm_*' and 'xen_' to a common prefix ('virt_*' ?).<br /> + This can be done by an automated script and verified by the unit tests.<br /> <br /> 3) Work gradually on code that would generate more impact, implementing abstractions and other mechanisms<br /> + The config file and thus the params passed to tests already add a lot of flexibility [see example snippet #1]<br /> + Improve monitor functionality, introducing methods that wrap the hypervisor instead of sending specific commands [see example snippet #2, #3]. <br /> + This would eventually become a rather large API. libvirt API should be evaluated and if considered sound and appropriate for autotest, could be made similar.<br /> + Modify tests so that they make use of improved Monitor/VM API<br /> <br /> 4) Publish resulting API and best practices documentation so that new tests writers may submit <br /> <br /> <br /> == Other notes ==<br /> <br /> * Regarding the prefix (virt_* ?), some functionality is not even tied to virtualization (such as the configuration format/parser) and could benefit other complex test modules;<br /> <br /> * kvm_vm:<br /> - There are significant differences in (KVM) VM class and (xen) XendDomain. Methods such as create() are quite different in parameters in implementation<br /> - (KVM) VM uses a 'monitor' to interact with qemu, whether (xen) XenDomain executes 'xm' commands <br /> - the 'monitor' abstraction is valid, but should have clear boundaries <br /> - xen autotest should also get a 'monitor' implementation<br /> - this even paves the way to a libvirt based monitor implementation<br /> - probably a capabilities mechanism would be necessary to acomodate diferences between virt technologies (see kvm_monitor below)<br /> <br /> * kvm_monitor<br /> - There are methods and parameters which do not map completely to Xen, specially PV guests (screendump, migrate with full disk copy, mouse move, etc)<br /> - Again, a capabilities mechanism could be used<br /> <br /> <br /> == Example #1 (pci_hotplug.py) ==<br /> <br /> Parameters provide a lot of flexibility and diminish the need to implement a interface on a variant that does not need or cannot provide it. This code snippet:<br /> <br /> 26 # Modprobe the module if specified in config file<br /> 27 module = params.get(&quot;modprobe_module&quot;)<br /> 28 if module:<br /> 29 session.cmd(&quot;modprobe %s&quot; % module)<br /> <br /> could be kept exactly like that, loading 'acpiphp' as it does now for RHEL guests onon kvm and nothing for RHEL xen domU kernels (with XEN_PCIDEV_FRONTEND=y). <br /> <br /> If someone wants to test a xen domU kernel with modular XEN_PCIDEV_FRONTEND, only this parameter would have to be adjusted.<br /> <br /> == Example #2 (pci_hotplug.py) ==<br /> <br /> Only operations from the monitor side (vm.monitor.*) would have to be implemented for each virtualization technology. On the guest side (session.cmd_output() and the like), things could go unchanged, helped by parameters when necessary.<br /> <br /> Curent code:<br /> <br /> 31 # Get output of command 'info pci' as reference<br /> 32 info_pci_ref = vm.monitor.info(&quot;pci&quot;)<br /> 33<br /> 34 # Get output of command as reference<br /> 35 reference = session.cmd_output(params.get(&quot;reference_cmd&quot;))<br /> <br /> Suggested code:<br /> <br /> 31 # Get output of command 'info pci' as reference<br /> 32 info_pci_ref = vm.monitor.get_pci_info()<br /> 33<br /> 34 # Get output of command as reference<br /> 35 reference = session.cmd_output(params.get(&quot;reference_cmd&quot;))<br /> <br /> <br /> == Example #3 (pci_hotplug.py) ==<br /> <br /> Current test code not only interacts directly with qemu but also deals with different versions and syntaxes:<br /> <br /> 41 # Probe qemu to verify what is the supported syntax for PCI hotplug<br /> 42 cmd_output = vm.monitor.cmd(&quot;?&quot;)<br /> 43 if len(re.findall(&quot;\ndevice_add&quot;, cmd_output)) &gt; 0:<br /> 44 cmd_type = &quot;device_add&quot;<br /> 45 elif len(re.findall(&quot;\npci_add&quot;, cmd_output)) &gt; 0:<br /> 46 cmd_type = &quot;pci_add&quot;<br /> 47 else:<br /> 48 raise error.TestError(&quot;Unknow version of qemu&quot;)<br /> <br /> Again, this should be abstracted in the monitor implementation, and consumed by the test code (this is even commented in the current code):<br /> <br /> 59 # Execute pci_add (should be replaced by a proper monitor method call)<br /> 60 add_output = vm.monitor.cmd(pci_add_cmd)<br /> 61 if not &quot;OK domain&quot; in add_output:<br /> 62 raise error.TestFail(&quot;Add PCI device failed. &quot;<br /> 63 &quot;Monitor command is: %s, Output: %r&quot; %<br /> 64 (pci_add_cmd, add_output))<br /> 65 after_add = vm.monitor.info(&quot;pci&quot;)<br /> <br /> This should become:<br /> <br /> 59 # Execute pci_add <br /> 60 if not vm.monitor_pci_add(model):<br /> 61 raise error.TestFail(&quot;Add PCI device failed.&quot;)<br /> 62 after_add = vm.monitor.get_pci_info()<br /> <br /> <br /> == Library code under client/tests/kvm ==<br /> <br /> The componentes are sorted by depencies, so that working down the list would solve all depencies, except when circular dependencies exist.<br /> <br /> Filename Impact Dependencies Spec Tied Xen<br /> -------- ------ ------------ ---- ---- ---<br /> kvm_config.py Negligible kvm_utils (logging) No No Yes<br /> kvm_subprocess.py Negligible none No No Yes<br /> scan_results.py Negligible none No No Yes<br /> cd_hash.py Negligible kvm_utils (logging) No No No<br /> rss_file_transfer.py Negligible none No No No<br /> <br /> kvm_monitor.py High kvm_utils (random_str) No Yes No[1]<br /> kvm_vm.py High kvm_utils No Yes Yes<br /> kvm_subprocess<br /> kvm_monitor<br /> rss_file_transfer<br /> <br /> kvm_scheduler.py Medium kvm_utils No[2] Yes[3] No <br /> kvm_preprocessing.py Medium kvm_vm No Yes[4] Yes<br /> kvm_utils<br /> kvm_subprocess<br /> kvm_monitor<br /> kvm_utils.py High kvm_subprocess No Yes[5] Yes<br /> <br /> kvm_test_utils.py High kvm_utils No Yes Yes<br /> kvm_vm<br /> kvm_subprocess<br /> scan_results<br /> <br /> Caption:<br /> -------<br /> - Spec: Specific for KVM (that is, not applicable to other virt technologies)<br /> - Tied: Currently tied to KVM in its current implementation<br /> - Xen: Currently used by xen autotest or has equivalent/derived implementation<br /> <br /> [1] - Xen autotest currently does not implement a monitor, but this analysis recommends one to be implemented<br /> [2] - Currently xen autotest does not have a control file that uses parallel tests, so it's low priority but could be useful<br /> [3] - Tied only to kvm.VM and vm.monitor, used as abstractions<br /> [4] - Tied only to kvm.VM and vm.monitor, used as abstractions<br /> [5] - There's both kvm specific code and code that should be shared between xen and kvm autotest</div>

Lmr https://linux-kvm.org/index.php?title=KVM_Autotest_Refactor_page&diff=3432 2010-12-08T01:09:11Z

<p>Lmr: </p> <hr /> <div>= KVM autotest refactor =<br /> <br /> == Suggested approach ==<br /> <br /> 1) Move code that has negligible impact [see table below] from 'client/tests/kvm' to 'client' ASAP<br /> + This would immediatelly benefit xen autotest, since it is frequently lagging behing features and fixes. One example is newer syntax in kvm_config.py that was not available in the xen autotest counterpart.<br /> <br /> 2) Synchronize that code (step 1) usage between kvm and xen autotest<br /> + Rewrite current tests, changing imports of modules from 'kvm_*' and 'xen_' to a common prefix ('virt_*' ?).<br /> + This can be done by an automated script and verified by the unit tests.<br /> <br /> 3) Work gradually on code that would generate more impact, implementing abstractions and other mechanisms<br /> + The config file and thus the params passed to tests already add a lot of flexibility [see example snippet #1]<br /> + Improve monitor functionality, introducing methods that wrap the hypervisor instead of sending specific commands [see example snippet #2, #3]. <br /> + This would eventually become a rather large API. libvirt API should be evaluated and if considered sound and appropriate for autotest, could be made similar.<br /> + Modify tests so that they make use of improved Monitor/VM API<br /> <br /> 4) Publish resulting API and best practices documentation so that new tests writers may submit <br /> <br /> <br /> == Other notes ==<br /> <br /> * Regarding the prefix (virt_* ?), some functionality is not even tied to virtualization (such as the configuration format/parser) and could benefit other complex test modules;<br /> <br /> * kvm_vm:<br /> - There are significant differences in (KVM) VM class and (xen) XendDomain. Methods such as create() are quite different in parameters in implementation<br /> - (KVM) VM uses a 'monitor' to interact with qemu, whether (xen) XenDomain executes 'xm' commands <br /> - the 'monitor' abstraction is valid, but should have clear boundaries <br /> - xen autotest should also get a 'monitor' implementation<br /> - this even paves the way to a libvirt based monitor implementation<br /> - probably a capabilities mechanism would be necessary to acomodate diferences between virt technologies (see kvm_monitor below)<br /> <br /> * kvm_monitor<br /> - There are methods and parameters which do not map completely to Xen, specially PV guests (screendump, migrate with full disk copy, mouse move, etc)<br /> - Again, a capabilities mechanism could be used<br /> <br /> <br /> == Example #1 (pci_hotplug.py) ==<br /> <br /> Parameters provide a lot of flexibility and diminish the need to implement a interface on a variant that does not need or cannot provide it. This code snippet:<br /> <br /> 26 # Modprobe the module if specified in config file<br /> 27 module = params.get(&quot;modprobe_module&quot;)<br /> 28 if module:<br /> 29 session.cmd(&quot;modprobe %s&quot; % module)<br /> <br /> could be kept exactly like that, loading 'acpiphp' as it does now for RHEL guests onon kvm and nothing for RHEL xen domU kernels (with XEN_PCIDEV_FRONTEND=y). <br /> <br /> If someone wants to test a xen domU kernel with modular XEN_PCIDEV_FRONTEND, only this parameter would have to be adjusted.<br /> <br /> <br /> == Example #2 (pci_hotplug.py) ==<br /> <br /> Only operations from the monitor side (vm.monitor.*) would have to be implemented for each virtualization technology. On the guest side (session.cmd_output() and the like), things could go unchanged, helped by parameters when necessary.<br /> <br /> Curent code:<br /> <br /> 31 # Get output of command 'info pci' as reference<br /> 32 info_pci_ref = vm.monitor.info(&quot;pci&quot;)<br /> 33<br /> 34 # Get output of command as reference<br /> 35 reference = session.cmd_output(params.get(&quot;reference_cmd&quot;))<br /> <br /> Suggested code:<br /> <br /> 31 # Get output of command 'info pci' as reference<br /> 32 info_pci_ref = vm.monitor.get_pci_info()<br /> 33<br /> 34 # Get output of command as reference<br /> 35 reference = session.cmd_output(params.get(&quot;reference_cmd&quot;))<br /> <br /> <br /> == Example #3 (pci_hotplug.py) ==<br /> <br /> Current test code not only interacts directly with qemu but also deals with different versions and syntaxes:<br /> <br /> 41 # Probe qemu to verify what is the supported syntax for PCI hotplug<br /> 42 cmd_output = vm.monitor.cmd(&quot;?&quot;)<br /> 43 if len(re.findall(&quot;\ndevice_add&quot;, cmd_output)) &gt; 0:<br /> 44 cmd_type = &quot;device_add&quot;<br /> 45 elif len(re.findall(&quot;\npci_add&quot;, cmd_output)) &gt; 0:<br /> 46 cmd_type = &quot;pci_add&quot;<br /> 47 else:<br /> 48 raise error.TestError(&quot;Unknow version of qemu&quot;)<br /> <br /> Again, this should be abstracted in the monitor implementation, and consumed by the test code (this is even commented in the current code):<br /> <br /> 59 # Execute pci_add (should be replaced by a proper monitor method call)<br /> 60 add_output = vm.monitor.cmd(pci_add_cmd)<br /> 61 if not &quot;OK domain&quot; in add_output:<br /> 62 raise error.TestFail(&quot;Add PCI device failed. &quot;<br /> 63 &quot;Monitor command is: %s, Output: %r&quot; %<br /> 64 (pci_add_cmd, add_output))<br /> 65 after_add = vm.monitor.info(&quot;pci&quot;)<br /> <br /> This should become:<br /> <br /> 59 # Execute pci_add <br /> 60 if not vm.monitor_pci_add(model):<br /> 61 raise error.TestFail(&quot;Add PCI device failed.&quot;)<br /> 62 after_add = vm.monitor.get_pci_info()<br /> <br /> <br /> == Library code under client/tests/kvm ==<br /> <br /> The componentes are sorted by depencies, so that working down the list would solve all depencies, except when circular dependencies exist.<br /> <br /> Filename Impact Dependencies Spec Tied Xen<br /> -------- ------ ------------ ---- ---- ---<br /> kmv_config.py Negligible kvm_utils (logging) No No Yes<br /> kvm_subprocess.py Negligible none No No Yes<br /> scan_results.py Negligible none No No Yes<br /> cd_hash.py Negligible kvm_utils (logging) No No No<br /> rss_file_transfer.py Negligible none No No No<br /> <br /> kvm_monitor.py High kvm_utils (random_str) No Yes No[1]<br /> kvm_vm.py High kvm_utils No Yes Yes<br /> kvm_subprocess<br /> kvm_monitor<br /> rss_file_transfer<br /> <br /> kvm_scheduler.py Medium kvm_utils No[2] Yes[3] No <br /> kvm_preprocessing.py Medium kvm_vm No Yes[4] Yes<br /> kvm_utils<br /> kvm_subprocess<br /> kvm_monitor<br /> kvm_utils.py High kvm_subprocess No Yes[5] Yes<br /> <br /> kvm_test_utils.py High kvm_utils No Yes Yes<br /> kvm_vm<br /> kvm_subprocess<br /> scan_results<br /> <br /> Legend:<br /> -------<br /> - Spec: Specific for KVM (that is, not applicable to other virt technologies)<br /> - Tied: Currently tied to KVM in its current implementation<br /> - Xen: Currently used by xen autotest or has equivalent/derived implementation<br /> <br /> [1] - Xen autotest currently does not implement a monitor, but this analysis recommends one to be implemented<br /> [2] - Currently xen autotest does not have a control file that uses parallel tests, so it's low priority but could be useful<br /> [3] - Tied only to kvm.VM and vm.monitor, used as abstractions<br /> [4] - Tied only to kvm.VM and vm.monitor, used as abstractions<br /> [5] - There's both kvm specific code and code that should be shared between xen and kvm autotest</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3431 2010-12-08T01:06:47Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> * Autotest refactor for usage with libvirt/xen. See more detauls at the [[KVM Autotest Refactor page]]<br /> <br /> Tasks:<br /> <br /> ** Move kvm autotest libraries to client/bin directory, so every client side test can use <br /> ** Review xen-autotest to see what are the differences between our vm class and theirs. Hopefully in the future, xen-autotest and libvirt-autotest will be only different vm implementations of our virt-autotest framework :)<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Multi host migration - Already have a 1st patchset available, working to rebase it against a cleanup patchset made by Jason<br /> <br /> * Hotplug tests during migration<br /> * Work on setting up the conmux server for our internal test grid<br /> * Research how to trigger jobs based on events rather than using the web interface setting recurrent jobs<br /> * Set up a qemu-block job, with kwolf's qemu-block tree<br /> * Set up a vhost job, with the upstream job (mst's trees for kernel and userspace)<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Done ===<br /> <br /> * Infrastructure for testing different vm storage backends: file/lvm/iSCSI - Done by Beijing team<br /> * Made some internal patches appropriate for upstream<br /> * Copy kvm-autotest@redhat.com on the email list of results<br /> * Send code to retrieve host kernels based on koji/brew tags upstream<br /> * Resolved UUID mount point check problem<br /> * QMP suite integrated upstream<br /> * Wrote code to retrieve kernels based on koji/brew tags<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> <br /> === done ===<br /> <br /> * Post-commit review of new network patches (committed on Oct 7 2010)<br /> * Use more exceptions in utility functions:<br /> ** Add convenience functions to kvm_subprocess.py to make running shell commands shorter<br /> ** Add convenience functions to the VM class<br /> * Refactor _get_command_output() and friends in kvm_monitor.py and add cmd_raw(), cmd_obj() and cmd_qmp() as required by Luiz Capitulino's test suite<br /> * Use select() instead of sleep() in kvm_monitor.py<br /> * Simplify migration code if possible (in kvm_vm.py and kvm_test_utils.py)<br /> * Fix VM.get_ifname() somehow (uses self.vnc_port -- bad)<br /> <br /> == akong ==<br /> <br /> * push redhat internal kvm subtests(18) to upstream<br /> ''trans_hugepage_defrag.py trans_hugepage.py trans_hugepage_swapping.py hdparm.py cpuflags.py multi_disk.py nmi_watchdog.py qmp_command.py qmp_event_notification.py cdrom.py fillup_disk.py floppy.py format_disk.py lvm.py module_probe.py stop_continue.py usb.py image_copy.py''<br /> <br /> * Add a new test: check if guest transmits packets when link is up/down<br /> * Report problem with multiple NICs and MAC address tracking [problem fixed]<br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge [request changed]<br /> * Bonding test [done]<br /> <br /> == jason ==<br /> <br /> * Write a 'cli' wrapper to make autotest handy/easier for developers<br /> * Tests in parallel with migration (reboot/unattended installation/file transfer/ autotest client tests / guest_script,autoit? / memtest86+(steps?)), need modifications to framework?<br /> * Clean the Migration tests: mainly for exec and add offline support<br /> * Refactor hotplug test using qdev and netdev_del/drive_del<br /> * Refactor the network cmdline generation, using peer when possible and vlan for elder qemu-kvm<br /> * Port network tests into windows<br /> * Migration to exec test suggested by mst<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3430 2010-12-07T16:59:04Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> * Autotest refactor for usage with libvirt/xen<br /> <br /> Tasks:<br /> <br /> ** Move kvm autotest libraries to client/bin directory, so every client side test can use <br /> ** Review xen-autotest to see what are the differences between our vm class and theirs. Hopefully in the future, xen-autotest and libvirt-autotest will be only different vm implementations of our virt-autotest framework :)<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Multi host migration - Already have a 1st patchset available, working to rebase it against a cleanup patchset made by Jason<br /> <br /> * Hotplug tests during migration<br /> * Work on setting up the conmux server for our internal test grid<br /> * Research how to trigger jobs based on events rather than using the web interface setting recurrent jobs<br /> * Set up a qemu-block job, with kwolf's qemu-block tree<br /> * Set up a vhost job, with the upstream job (mst's trees for kernel and userspace)<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Done ===<br /> <br /> * Infrastructure for testing different vm storage backends: file/lvm/iSCSI - Done by Beijing team<br /> * Made some internal patches appropriate for upstream<br /> * Copy kvm-autotest@redhat.com on the email list of results<br /> * Send code to retrieve host kernels based on koji/brew tags upstream<br /> * Resolved UUID mount point check problem<br /> * QMP suite integrated upstream<br /> * Wrote code to retrieve kernels based on koji/brew tags<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> <br /> === done ===<br /> <br /> * Post-commit review of new network patches (committed on Oct 7 2010)<br /> * Use more exceptions in utility functions:<br /> ** Add convenience functions to kvm_subprocess.py to make running shell commands shorter<br /> ** Add convenience functions to the VM class<br /> * Refactor _get_command_output() and friends in kvm_monitor.py and add cmd_raw(), cmd_obj() and cmd_qmp() as required by Luiz Capitulino's test suite<br /> * Use select() instead of sleep() in kvm_monitor.py<br /> * Simplify migration code if possible (in kvm_vm.py and kvm_test_utils.py)<br /> * Fix VM.get_ifname() somehow (uses self.vnc_port -- bad)<br /> <br /> == akong ==<br /> <br /> * push redhat internal kvm subtests(18) to upstream<br /> ''trans_hugepage_defrag.py trans_hugepage.py trans_hugepage_swapping.py hdparm.py cpuflags.py multi_disk.py nmi_watchdog.py qmp_command.py qmp_event_notification.py cdrom.py fillup_disk.py floppy.py format_disk.py lvm.py module_probe.py stop_continue.py usb.py image_copy.py''<br /> <br /> * Add a new test: check if guest transmits packets when link is up/down<br /> * Report problem with multiple NICs and MAC address tracking [problem fixed]<br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge [request changed]<br /> * Bonding test [done]<br /> <br /> == jason ==<br /> <br /> * Write a 'cli' wrapper to make autotest handy/easier for developers<br /> * Tests in parallel with migration (reboot/unattended installation/file transfer/ autotest client tests / guest_script,autoit? / memtest86+(steps?)), need modifications to framework?<br /> * Clean the Migration tests: mainly for exec and add offline support<br /> * Refactor hotplug test using qdev and netdev_del/drive_del<br /> * Refactor the network cmdline generation, using peer when possible and vlan for elder qemu-kvm<br /> * Port network tests into windows<br /> * Migration to exec test suggested by mst<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3424 2010-12-01T11:56:57Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> * Autotest refactor for usage with libvirt/xen<br /> <br /> Tasks:<br /> <br /> ** Move kvm autotest libraries to client/bin directory, so every client side test can use <br /> ** Review xen-autotest to see what are the differences between our vm class and theirs. Hopefully in the future, xen-autotest and libvirt-autotest will be only different vm implementations of our virt-autotest framework :)<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Multi host migration - priority increased<br /> <br /> * Infrastructure for testing different vm storage backends: file/lvm/iSCSI<br /> * Work on setting up the conmux server for our internal test grid<br /> * Research how to trigger jobs based on events rather than using the web interface setting recurrent jobs<br /> * Set up a qemu-block job, with kwolf's qemu-block tree<br /> * Set up a vhost job, with the upstream job (mst's trees for kernel and userspace)<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Done ===<br /> <br /> * Made some internal patches appropriate for upstream<br /> * Copy kvm-autotest@redhat.com on the email list of results<br /> * Send code to retrieve host kernels based on koji/brew tags upstream<br /> * Resolved UUID mount point check problem<br /> * QMP suite integrated upstream<br /> * Wrote code to retrieve kernels based on koji/brew tags<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> <br /> === done ===<br /> <br /> * Post-commit review of new network patches (committed on Oct 7 2010)<br /> * Use more exceptions in utility functions:<br /> ** Add convenience functions to kvm_subprocess.py to make running shell commands shorter<br /> ** Add convenience functions to the VM class<br /> * Refactor _get_command_output() and friends in kvm_monitor.py and add cmd_raw(), cmd_obj() and cmd_qmp() as required by Luiz Capitulino's test suite<br /> * Use select() instead of sleep() in kvm_monitor.py<br /> * Simplify migration code if possible (in kvm_vm.py and kvm_test_utils.py)<br /> * Fix VM.get_ifname() somehow (uses self.vnc_port -- bad)<br /> <br /> == akong ==<br /> <br /> * push redhat internal kvm subtests(18) to upstream<br /> ''trans_hugepage_defrag.py trans_hugepage.py trans_hugepage_swapping.py hdparm.py cpuflags.py multi_disk.py nmi_watchdog.py qmp_command.py qmp_event_notification.py cdrom.py fillup_disk.py floppy.py format_disk.py lvm.py module_probe.py stop_continue.py usb.py image_copy.py''<br /> <br /> * Add a new test: check if guest transmits packets when link is up/down<br /> * Report problem with multiple NICs and MAC address tracking [problem fixed]<br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge [request changed]<br /> * Bonding test [done]<br /> <br /> == jason ==<br /> <br /> * Write a 'cli' wrapper to make autotest handy/easier for developers<br /> * Tests in parallel with migration (reboot/unattended installation/file transfer/ autotest client tests / guest_script,autoit? / memtest86+(steps?)), need modifications to framework?<br /> * Clean the Migration tests: mainly for exec and add offline support<br /> * Refactor hotplug test using qdev and netdev_del/drive_del<br /> * Refactor the network cmdline generation, using peer when possible and vlan for elder qemu-kvm<br /> * Port network tests into windows<br /> * Migration to exec test suggested by mst<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3423 2010-12-01T11:55:57Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> * Autotest refactor for usage with libvirt/xen<br /> Tasks:<br /> * Move kvm autotest libraries to client/bin directory, so every client side test can use <br /> * Review xen-autotest to see what are the differences between our vm class and theirs. Hopefully in the future, xen-autotest and libvirt-autotest will be only different vm implementations of our virt-autotest framework :)<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Multi host migration - priority increased<br /> <br /> * Infrastructure for testing different vm storage backends: file/lvm/iSCSI<br /> * Work on setting up the conmux server for our internal test grid<br /> * Research how to trigger jobs based on events rather than using the web interface setting recurrent jobs<br /> * Set up a qemu-block job, with kwolf's qemu-block tree<br /> * Set up a vhost job, with the upstream job (mst's trees for kernel and userspace)<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Done ===<br /> <br /> * Made some internal patches appropriate for upstream<br /> * Copy kvm-autotest@redhat.com on the email list of results<br /> * Send code to retrieve host kernels based on koji/brew tags upstream<br /> * Resolved UUID mount point check problem<br /> * QMP suite integrated upstream<br /> * Wrote code to retrieve kernels based on koji/brew tags<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> <br /> === done ===<br /> <br /> * Post-commit review of new network patches (committed on Oct 7 2010)<br /> * Use more exceptions in utility functions:<br /> ** Add convenience functions to kvm_subprocess.py to make running shell commands shorter<br /> ** Add convenience functions to the VM class<br /> * Refactor _get_command_output() and friends in kvm_monitor.py and add cmd_raw(), cmd_obj() and cmd_qmp() as required by Luiz Capitulino's test suite<br /> * Use select() instead of sleep() in kvm_monitor.py<br /> * Simplify migration code if possible (in kvm_vm.py and kvm_test_utils.py)<br /> * Fix VM.get_ifname() somehow (uses self.vnc_port -- bad)<br /> <br /> == akong ==<br /> <br /> * push redhat internal kvm subtests(18) to upstream<br /> ''trans_hugepage_defrag.py trans_hugepage.py trans_hugepage_swapping.py hdparm.py cpuflags.py multi_disk.py nmi_watchdog.py qmp_command.py qmp_event_notification.py cdrom.py fillup_disk.py floppy.py format_disk.py lvm.py module_probe.py stop_continue.py usb.py image_copy.py''<br /> <br /> * Add a new test: check if guest transmits packets when link is up/down<br /> * Report problem with multiple NICs and MAC address tracking [problem fixed]<br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge [request changed]<br /> * Bonding test [done]<br /> <br /> == jason ==<br /> <br /> * Write a 'cli' wrapper to make autotest handy/easier for developers<br /> * Tests in parallel with migration (reboot/unattended installation/file transfer/ autotest client tests / guest_script,autoit? / memtest86+(steps?)), need modifications to framework?<br /> * Clean the Migration tests: mainly for exec and add offline support<br /> * Refactor hotplug test using qdev and netdev_del/drive_del<br /> * Refactor the network cmdline generation, using peer when possible and vlan for elder qemu-kvm<br /> * Port network tests into windows<br /> * Migration to exec test suggested by mst<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3405 2010-11-24T04:02:16Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> * Autotest refactor for usage with libvirt/xen<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Multi host migration - priority increased<br /> <br /> * Infrastructure for testing different vm storage backends: file/lvm/iSCSI<br /> * Work on setting up the conmux server for our internal test grid<br /> * Research how to trigger jobs based on events rather than using the web interface setting recurrent jobs<br /> * Set up a qemu-block job, with kwolf's qemu-block tree<br /> * Set up a vhost job, with the upstream job (mst's trees for kernel and userspace)<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Done ===<br /> <br /> * Made some internal patches appropriate for upstream<br /> * Copy kvm-autotest@redhat.com on the email list of results<br /> * Send code to retrieve host kernels based on koji/brew tags upstream<br /> * Resolved UUID mount point check problem<br /> * QMP suite integrated upstream<br /> * Wrote code to retrieve kernels based on koji/brew tags<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> <br /> === done ===<br /> <br /> * Post-commit review of new network patches (committed on Oct 7 2010)<br /> * Use more exceptions in utility functions:<br /> ** Add convenience functions to kvm_subprocess.py to make running shell commands shorter<br /> ** Add convenience functions to the VM class<br /> * Refactor _get_command_output() and friends in kvm_monitor.py and add cmd_raw(), cmd_obj() and cmd_qmp() as required by Luiz Capitulino's test suite<br /> * Use select() instead of sleep() in kvm_monitor.py<br /> * Simplify migration code if possible (in kvm_vm.py and kvm_test_utils.py)<br /> * Fix VM.get_ifname() somehow (uses self.vnc_port -- bad)<br /> <br /> == akong ==<br /> <br /> * Report problem with multiple NICs and MAC address tracking<br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge<br /> * Bonding test<br /> * Add a new test: check if guest transmits packets when link is up/down<br /> <br /> == jason ==<br /> <br /> * Write a 'cli' wrapper to make autotest handy/easier for developers<br /> * Tests in parallel with migration (reboot/unattended installation/file transfer/ autotest client tests / guest_script,autoit? / memtest86+(steps?)), need modifications to framework?<br /> * Clean the Migration tests: mainly for exec and add offline support<br /> * Refactor hotplug test using qdev and netdev_del/drive_del<br /> * Refactor the network cmdline generation, using peer when possible and vlan for elder qemu-kvm<br /> * Port network tests into windows<br /> * Migration to exec test suggested by mst<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3399 2010-11-17T11:20:08Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> * Autotest refactor for usage with libvirt/xen<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Multi host migration - priority increased<br /> <br /> * Copy kvm-autotest@redhat.com on the email list of results<br /> * Infrastructure for testing different vm storage backends: file/lvm/iSCSI<br /> * Work on setting up the conmux server for our internal test grid<br /> * Research how to trigger jobs based on events rather than using the web interface setting recurrent jobs<br /> * Set up a qemu-block job, with kwolf's qemu-block tree<br /> * Set up a vhost job, with the upstream job (mst's trees for kernel and userspace)<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Done ===<br /> <br /> * Send code to retrieve host kernels based on koji/brew tags upstream<br /> * Resolved UUID mount point check problem<br /> * QMP suite integrated upstream<br /> * Wrote code to retrieve kernels based on koji/brew tags<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> * Post-commit review of new network patches (committed on Oct 7 2010)<br /> * Use more exceptions in utility functions:<br /> ** Add convenience functions to kvm_subprocess.py to make running shell commands shorter<br /> ** Add convenience functions to the VM class<br /> * Refactor _get_command_output() and friends in kvm_monitor.py and add cmd_raw(), cmd_obj() and cmd_qmp() as required by Luiz Capitulino's test suite<br /> * Use select() instead of sleep() in kvm_monitor.py<br /> * Simplify migration code if possible (in kvm_vm.py and kvm_test_utils.py)<br /> * Fix VM.get_ifname() somehow (uses self.vnc_port -- bad)<br /> <br /> == akong ==<br /> <br /> * Report problem with multiple NICs and MAC address tracking<br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge<br /> * Bonding test<br /> * Add a new test: check if guest transmits packets when link is up/down<br /> <br /> == jason ==<br /> <br /> * Write a 'cli' wrapper to make autotest handy/easier for developers<br /> * Tests in parallel with migration (reboot/unattended installation/file transfer/ autotest client tests / guest_script,autoit? / memtest86+(steps?)), need modifications to framework?<br /> * Refactor hotplug test using qdev and netdev_del/drive_del<br /> * Refactor the network cmdline generation, using peer when possible and vlan for elder qemu-kvm<br /> * Port network tests into windows<br /> * Migration to exec test suggested by mst<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3398 2010-11-17T10:36:24Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> * Autotest refactor for usage with libvirt/xen<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Multi host migration - priority increased<br /> <br /> * Infrastructure for testing different vm storage backends: file/lvm/iSCSI<br /> * Work on setting up the conmux server for our internal test grid<br /> * Research how to trigger jobs based on events rather than using the web interface setting recurrent jobs<br /> * Set up a qemu-block job, with kwolf's qemu-block tree<br /> * Set up a vhost job, with the upstream job (mst's trees for kernel and userspace)<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Done ===<br /> <br /> * Send code to retrieve host kernels based on koji/brew tags upstream<br /> * Resolved UUID mount point check problem<br /> * QMP suite integrated upstream<br /> * Wrote code to retrieve kernels based on koji/brew tags<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> * Post-commit review of new network patches (committed on Oct 7 2010)<br /> * Use more exceptions in utility functions:<br /> ** Add convenience functions to kvm_subprocess.py to make running shell commands shorter<br /> ** Add convenience functions to the VM class<br /> * Refactor _get_command_output() and friends in kvm_monitor.py and add cmd_raw(), cmd_obj() and cmd_qmp() as required by Luiz Capitulino's test suite<br /> * Use select() instead of sleep() in kvm_monitor.py<br /> * Simplify migration code if possible (in kvm_vm.py and kvm_test_utils.py)<br /> * Fix VM.get_ifname() somehow (uses self.vnc_port -- bad)<br /> <br /> == akong ==<br /> <br /> * Report problem with multiple NICs and MAC address tracking<br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge<br /> * Bonding test<br /> * Add a new test: check if guest transmits packets when link is up/down<br /> <br /> == jason ==<br /> <br /> * Write a 'cli' wrapper to make autotest handy/easier for developers<br /> * Tests in parallel with migration (reboot/unattended installation/file transfer/ autotest client tests / guest_script,autoit? / memtest86+(steps?)), need modifications to framework?<br /> * Refactor hotplug test using qdev and netdev_del/drive_del<br /> * Refactor the network cmdline generation, using peer when possible and vlan for elder qemu-kvm<br /> * Port network tests into windows<br /> * Migration to exec test suggested by mst<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3397 2010-11-17T10:34:03Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> * Autotest refactor for usage with libvirt/xen<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Multi host migration - priority increased<br /> <br /> * Infrastructure for testing different vm storage backends: file/lvm/iSCSI<br /> * Work on setting up the conmux server for our internal test grid<br /> * Research how to trigger jobs based on events rather than using the web interface setting recurrent jobs<br /> * Set up a qemu-block job, with kwolf's qemu-block tree<br /> * Set up a vhost job, with the upstream job (mst's trees for kernel and userspace)<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Done ===<br /> <br /> * Send code to retrieve host kernels based on koji/brew tags upstream<br /> * Resolved UUID mount point check problem<br /> * QMP suite integrated upstream<br /> * Wrote code to retrieve kernels based on koji/brew tags<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> * Post-commit review of new network patches (committed on Oct 7 2010)<br /> * Use more exceptions in utility functions:<br /> ** Add convenience functions to kvm_subprocess.py to make running shell commands shorter<br /> ** Add convenience functions to the VM class<br /> * Refactor _get_command_output() and friends in kvm_monitor.py and add cmd_raw(), cmd_obj() and cmd_qmp() as required by Luiz Capitulino's test suite<br /> * Use select() instead of sleep() in kvm_monitor.py<br /> * Simplify migration code if possible (in kvm_vm.py and kvm_test_utils.py)<br /> * Fix VM.get_ifname() somehow (uses self.vnc_port -- bad)<br /> <br /> == akong ==<br /> <br /> * Report problem with multiple NICs and MAC address tracking<br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge<br /> * Bonding test<br /> * Add a new test: check if guest transmits packets when link is up/down<br /> <br /> == jason ==<br /> <br /> * Tests in parallel with migration (reboot/unattended installation/file transfer/ autotest client tests / guest_script,autoit? / memtest86+(steps?)), need modifications to framework?<br /> * Refactor hotplug test using qdev and netdev_del/drive_del<br /> * Refactor the network cmdline generation, using peer when possible and vlan for elder qemu-kvm<br /> * Port network tests into windows<br /> * Migration to exec test suggested by mst<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3396 2010-11-17T01:15:57Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> * Libvirt integration - Will hold on this work item for the next couple of months due to resource constraints.<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Multi host migration - priority increased<br /> <br /> * Infrastructure for testing different vm storage backends: file/lvm/iSCSI<br /> * Work on setting up the conmux server for our internal test grid<br /> * Research how to trigger jobs based on events rather than using the web interface setting recurrent jobs<br /> * Set up a qemu-block job, with kwolf's qemu-block tree<br /> * Set up a vhost job, with the upstream job (mst's trees for kernel and userspace)<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Done ===<br /> <br /> * Send code to retrieve host kernels based on koji/brew tags upstream<br /> * Resolved UUID mount point check problem<br /> * QMP suite integrated upstream<br /> * Wrote code to retrieve kernels based on koji/brew tags<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> * Post-commit review of new network patches (committed on Oct 7 2010)<br /> * Use more exceptions in utility functions:<br /> ** Add convenience functions to kvm_subprocess.py to make running shell commands shorter<br /> ** Add convenience functions to the VM class<br /> * Refactor _get_command_output() and friends in kvm_monitor.py and add cmd_raw(), cmd_obj() and cmd_qmp() as required by Luiz Capitulino's test suite<br /> * Use select() instead of sleep() in kvm_monitor.py<br /> * Simplify migration code if possible (in kvm_vm.py and kvm_test_utils.py)<br /> * Fix VM.get_ifname() somehow (uses self.vnc_port -- bad)<br /> <br /> == akong ==<br /> <br /> * Report problem with multiple NICs and MAC address tracking<br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge<br /> * Bonding test<br /> * Add a new test: check if guest transmits packets when link is up/down<br /> <br /> == jason ==<br /> <br /> * Tests in parallel with migration (reboot/unattended installation/file transfer/ autotest client tests / guest_script,autoit? / memtest86+(steps?)), need modifications to framework?<br /> * Refactor hotplug test using qdev and netdev_del/drive_del<br /> * Refactor the network cmdline generation, using peer when possible and vlan for elder qemu-kvm<br /> * Port network tests into windows<br /> * Migration to exec test suggested by mst<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3383 2010-11-04T11:21:16Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> * Libvirt integration - Will hold on this work item for the next couple of months due to resource constraints.<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Multi host migration - priority increased<br /> <br /> * Infrastructure for testing different vm storage backends: file/lvm/iSCSI<br /> * Work on setting up the conmux server for our internal test grid<br /> * Research how to trigger jobs based on events rather than using the web interface setting recurrent jobs<br /> * Set up a qemu-block job, with kwolf's qemu-block tree<br /> * Set up a vhost job, with the upstream job (mst's trees for kernel and userspace)<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Done ===<br /> <br /> * Send code to retrieve host kernels based on koji/brew tags upstream<br /> * Resolved UUID mount point check problem<br /> * QMP suite integrated upstream<br /> * Wrote code to retrieve kernels based on koji/brew tags<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> * Post-commit review of new network patches (committed on Oct 7 2010)<br /> * Use more exceptions in utility functions:<br /> ** Add convenience functions to kvm_subprocess.py to make running shell commands shorter<br /> ** Add convenience functions to the VM class<br /> * Refactor _get_command_output() and friends in kvm_monitor.py and add cmd_raw(), cmd_obj() and cmd_qmp() as required by Luiz Capitulino's test suite<br /> * Use select() instead of sleep() in kvm_monitor.py<br /> * Simplify migration code if possible (in kvm_vm.py and kvm_test_utils.py)<br /> * Fix VM.get_ifname() somehow (uses self.vnc_port -- bad)<br /> <br /> == akong ==<br /> <br /> * Report problem with multiple NICs and MAC address tracking<br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge<br /> * Bonding test<br /> * Add a new test: check if guest transmits packets when link is up/down<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3368 2010-10-28T10:13:48Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> * Libvirt integration - Will hold on this work item for the next couple of months due to resource constraints.<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Multi host migration - priority increased<br /> <br /> * Work on setting up the conmux server for our internal test grid<br /> * Research how to trigger jobs based on events rather than using the web interface setting recurrent jobs<br /> * Set up a qemu-block job, with kwolf's qemu-block tree<br /> * Set up a vhost job, with the upstream job (mst's trees for kernel and userspace)<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Done ===<br /> <br /> * Send code to retrieve host kernels based on koji/brew tags upstream<br /> * Resolved UUID mount point check problem<br /> * QMP suite integrated upstream<br /> * Wrote code to retrieve kernels based on koji/brew tags<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> * Post-commit review of new network patches (committed on Oct 7 2010)<br /> * Use more exceptions in utility functions:<br /> ** Add convenience functions to kvm_subprocess.py to make running shell commands shorter<br /> ** Add convenience functions to the VM class<br /> * Refactor _get_command_output() and friends in kvm_monitor.py and add cmd_raw(), cmd_obj() and cmd_qmp() as required by Luiz Capitulino's test suite<br /> * Use select() instead of sleep() in kvm_monitor.py<br /> * Simplify migration code if possible (in kvm_vm.py and kvm_test_utils.py)<br /> * Fix VM.get_ifname() somehow (uses self.vnc_port -- bad)<br /> <br /> == akong ==<br /> <br /> * Report problem with multiple NICs and MAC address tracking<br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge<br /> * Bonding test<br /> * Add a new test: check if guest transmits packets when link is up/down<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3362 2010-10-21T10:21:06Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> * Libvirt integration - Will hold on this work item for the next couple of months due to resource constraints.<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Multi host migration - priority increased<br /> <br /> * Research how to trigger jobs based on events rather than using the web interface setting recurrent jobs<br /> * Set up a qemu-block job, with kwolf's qemu-block tree<br /> * Set up a vhost job, with the upstream job (mst's trees for kernel and userspace)<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Done ===<br /> <br /> * Send code to retrieve host kernels based on koji/brew tags upstream<br /> * Resolved UUID mount point check problem<br /> * QMP suite integrated upstream<br /> * Wrote code to retrieve kernels based on koji/brew tags<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> * Post-commit review of new network patches (committed on Oct 7 2010)<br /> * Use more exceptions in utility functions:<br /> ** Add convenience functions to kvm_subprocess.py to make running shell commands shorter<br /> ** Add convenience functions to the VM class<br /> * Refactor _get_command_output() and friends in kvm_monitor.py and add cmd_raw(), cmd_obj() and cmd_qmp() as required by Luiz Capitulino's test suite<br /> <br /> == akong ==<br /> <br /> * Report problem with multiple NICs and MAC address tracking<br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge<br /> * Bonding test<br /> * Add a new test: check if guest transmits packets when link is up/down<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3361 2010-10-21T10:07:31Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> * Libvirt integration - Will hold on this work item for the next couple of months due to resource constraints.<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Multi host migration - priority increased<br /> <br /> * Research how to trigger jobs based on events rather than using the web interface setting recurrent jobs<br /> * Set up a qemu-block job, with kwolf's qemu-block tree<br /> * Set up a vhost job, with the upstream job (mst's trees for kernel and userspace)<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Done ===<br /> <br /> * Send code to retrieve host kernels based on koji/brew tags upstream<br /> * Resolved UUID mount point check problem<br /> <br /> * Write code to retrieve kernels based on koji/brew tags<br /> * Set up new autotest server<br /> * Resolved autotest bugs that prevented proper kernel build/install functionality for Fedora/newer RHEL systems:<br /> http://patchwork.test.kernel.org/patch/2630/<br /> http://patchwork.test.kernel.org/patch/2631/<br /> http://patchwork.test.kernel.org/patch/2632/<br /> http://patchwork.test.kernel.org/patch/2640/<br /> http://patchwork.test.kernel.org/patch/2634/<br /> http://patchwork.test.kernel.org/patch/2635/<br /> <br /> Now the host kernel install control file patch is fully functional:<br /> http://patchwork.test.kernel.org/patch/2641/<br /> <br /> * Made cache=off the default for KVM autotest tests:<br /> http://patchwork.test.kernel.org/patch/2642/<br /> * Removed boot=on when booting virtio images:<br /> http://patchwork.test.kernel.org/patch/2643/<br /> <br /> * Applied network patchset (Yay!)<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> * Post-commit review of new network patches (committed on Oct 7 2010)<br /> * Use more exceptions in utility functions:<br /> ** Add convenience functions to kvm_subprocess.py to make running shell commands shorter<br /> ** Add convenience functions to the VM class<br /> * Refactor _get_command_output() and friends in kvm_monitor.py and add cmd_raw(), cmd_obj() and cmd_qmp() as required by Luiz Capitulino's test suite<br /> <br /> == akong ==<br /> <br /> * Report problem with multiple NICs and MAC address tracking<br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge<br /> * Bonding test<br /> * Add a new test: check if guest transmits packets when link is up/down<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3357 2010-10-15T11:46:57Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> * Libvirt integration - Will hold on this work item for the next couple of months due to resource constraints.<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Multi host migration - priority increased<br /> <br /> * Send code to retrieve host kernels based on koji/brew tags upstream<br /> * Research how to trigger jobs based on events rather than using the web interface setting recurrent jobs<br /> * Set up a qemu-block job, with kwolf's qemu-block tree<br /> * Set up a vhost job, with the upstream job (mst's trees for kernel and userspace)<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Done ===<br /> <br /> * Write code to retrieve kernels based on koji/brew tags<br /> * Set up new autotest server<br /> * Resolved autotest bugs that prevented proper kernel build/install functionality for Fedora/newer RHEL systems:<br /> http://patchwork.test.kernel.org/patch/2630/<br /> http://patchwork.test.kernel.org/patch/2631/<br /> http://patchwork.test.kernel.org/patch/2632/<br /> http://patchwork.test.kernel.org/patch/2640/<br /> http://patchwork.test.kernel.org/patch/2634/<br /> http://patchwork.test.kernel.org/patch/2635/<br /> <br /> Now the host kernel install control file patch is fully functional:<br /> http://patchwork.test.kernel.org/patch/2641/<br /> <br /> * Made cache=off the default for KVM autotest tests:<br /> http://patchwork.test.kernel.org/patch/2642/<br /> * Removed boot=on when booting virtio images:<br /> http://patchwork.test.kernel.org/patch/2643/<br /> <br /> * Applied network patchset (Yay!)<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> * Post-commit review of new network patches (committed on Oct 7 2010)<br /> * Use more exceptions in utility functions:<br /> ** Add convenience functions to kvm_subprocess.py to make running shell commands shorter<br /> ** Add convenience functions to the VM class<br /> <br /> == akong ==<br /> <br /> * Report problem with multiple NICs and MAC address tracking<br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge<br /> * Bonding test<br /> * Add a new test: check if guest transmits packets when link is up/down<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3356 2010-10-15T11:45:56Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> * Libvirt integration - Will hold on this work item for the next couple of months due to resource constraints.<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Multi host migration - priority increased<br /> <br /> * Research how to trigger jobs based on events rather than using the web interface setting recurrent jobs<br /> * Set up a qemu-block job, with kwolf's qemu-block tree<br /> * Set up a vhost job, with the upstream job (mst's trees for kernel and userspace)<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Done ===<br /> <br /> * Write code to retrieve kernels based on koji/brew tags<br /> * Set up new autotest server<br /> * Resolved autotest bugs that prevented proper kernel build/install functionality for Fedora/newer RHEL systems:<br /> http://patchwork.test.kernel.org/patch/2630/<br /> http://patchwork.test.kernel.org/patch/2631/<br /> http://patchwork.test.kernel.org/patch/2632/<br /> http://patchwork.test.kernel.org/patch/2640/<br /> http://patchwork.test.kernel.org/patch/2634/<br /> http://patchwork.test.kernel.org/patch/2635/<br /> <br /> Now the host kernel install control file patch is fully functional:<br /> http://patchwork.test.kernel.org/patch/2641/<br /> <br /> * Made cache=off the default for KVM autotest tests:<br /> http://patchwork.test.kernel.org/patch/2642/<br /> * Removed boot=on when booting virtio images:<br /> http://patchwork.test.kernel.org/patch/2643/<br /> <br /> * Applied network patchset (Yay!)<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> * Post-commit review of new network patches (committed on Oct 7 2010)<br /> * Use more exceptions in utility functions:<br /> ** Add convenience functions to kvm_subprocess.py to make running shell commands shorter<br /> ** Add convenience functions to the VM class<br /> <br /> == akong ==<br /> <br /> * Report problem with multiple NICs and MAC address tracking<br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge<br /> * Bonding test<br /> * Add a new test: check if guest transmits packets when link is up/down<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3353 2010-10-14T21:41:35Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> * Libvirt integration - Will hold on this work item for the next couple of months due to resource constraints.<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Multi host migration - priority increased<br /> <br /> * Write code to retrieve kernels based on koji/brew tags<br /> * Research how to trigger jobs based on events rather than using the web interface<br /> * Set up a qemu-block job, with kwolf's qemu-block tree<br /> * Set up a vhost job, with the upstream job (mst's trees for kernel and userspace)<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Done ===<br /> <br /> * Finish setting up new autotest server<br /> * Resolved autotest bugs that prevented proper kernel build/install functionality for Fedora/newer RHEL systems:<br /> http://patchwork.test.kernel.org/patch/2630/<br /> http://patchwork.test.kernel.org/patch/2631/<br /> http://patchwork.test.kernel.org/patch/2632/<br /> http://patchwork.test.kernel.org/patch/2640/<br /> http://patchwork.test.kernel.org/patch/2634/<br /> http://patchwork.test.kernel.org/patch/2635/<br /> <br /> Now the host kernel install control file patch is fully functional:<br /> http://patchwork.test.kernel.org/patch/2641/<br /> <br /> * Made cache=off the default for KVM autotest tests:<br /> http://patchwork.test.kernel.org/patch/2642/<br /> * Removed boot=on when booting virtio images:<br /> http://patchwork.test.kernel.org/patch/2643/<br /> <br /> * Applied network patchset (Yay!)<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> * Post-commit review of new network patches (committed on Oct 7 2010)<br /> * Use more exceptions in utility functions:<br /> ** Add convenience functions to kvm_subprocess.py to make running shell commands shorter<br /> ** Add convenience functions to the VM class<br /> <br /> == akong ==<br /> <br /> * Report problem with multiple NICs and MAC address tracking<br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge<br /> * Bonding test<br /> * Add a new test: check if guest transmits packets when link is up/down<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3352 2010-10-14T21:40:49Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> * Libvirt integration - Will hold on this work item for the next couple of months due to resource constraints.<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Multi host migration - priority increased<br /> <br /> * Research how to trigger jobs based on events rather than using the web interface<br /> * Set up a qemu-block job, with kwolf's qemu-block tree<br /> * Set up a vhost job, with the upstream job (mst's trees for kernel and userspace)<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Done ===<br /> <br /> * Fix bug http://autotest.kernel.org/ticket/42<br /> * Finish setting up new autotest server<br /> * Resolved autotest bugs that prevented proper kernel build/install functionality for Fedora/newer RHEL systems:<br /> http://patchwork.test.kernel.org/patch/2630/<br /> http://patchwork.test.kernel.org/patch/2631/<br /> http://patchwork.test.kernel.org/patch/2632/<br /> http://patchwork.test.kernel.org/patch/2640/<br /> http://patchwork.test.kernel.org/patch/2634/<br /> http://patchwork.test.kernel.org/patch/2635/<br /> <br /> Now the host kernel install control file patch is fully functional:<br /> http://patchwork.test.kernel.org/patch/2641/<br /> <br /> * Made cache=off the default for KVM autotest tests:<br /> http://patchwork.test.kernel.org/patch/2642/<br /> * Removed boot=on when booting virtio images:<br /> http://patchwork.test.kernel.org/patch/2643/<br /> <br /> * Applied network patchset (Yay!)<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> * Post-commit review of new network patches (committed on Oct 7 2010)<br /> * Use more exceptions in utility functions:<br /> ** Add convenience functions to kvm_subprocess.py to make running shell commands shorter<br /> ** Add convenience functions to the VM class<br /> <br /> == akong ==<br /> <br /> * Report problem with multiple NICs and MAC address tracking<br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge<br /> * Bonding test<br /> * Add a new test: check if guest transmits packets when link is up/down<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3348 2010-10-14T12:41:35Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> * Libvirt integration - Will hold on this work item for the next couple of months due to resource constraints.<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Multi host migration - priority increased<br /> * Fix bug http://autotest.kernel.org/ticket/42<br /> * Finish setting up new autotest server<br /> <br /> * Set up a qemu-block job, with kwolf's qemu-block tree<br /> * Set up a vhost job, with the upstream job (mst's trees for kernel and userspace)<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Done ===<br /> <br /> * Resolved autotest bugs that prevented proper kernel build/install functionality for Fedora/newer RHEL systems:<br /> http://patchwork.test.kernel.org/patch/2630/<br /> http://patchwork.test.kernel.org/patch/2631/<br /> http://patchwork.test.kernel.org/patch/2632/<br /> http://patchwork.test.kernel.org/patch/2640/<br /> http://patchwork.test.kernel.org/patch/2634/<br /> http://patchwork.test.kernel.org/patch/2635/<br /> <br /> Now the host kernel install control file patch is fully functional:<br /> http://patchwork.test.kernel.org/patch/2641/<br /> <br /> * Made cache=off the default for KVM autotest tests:<br /> http://patchwork.test.kernel.org/patch/2642/<br /> * Removed boot=on when booting virtio images:<br /> http://patchwork.test.kernel.org/patch/2643/<br /> <br /> * Applied network patchset (Yay!)<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> * Post-commit review of new network patches (committed on Oct 7 2010)<br /> * Use more exceptions in utility functions:<br /> ** Add convenience functions to kvm_subprocess.py to make running shell commands shorter<br /> ** Add convenience functions to the VM class<br /> <br /> == akong ==<br /> <br /> * Report problem with multiple NICs and MAC address tracking<br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge<br /> * Bonding test<br /> * Add a new test: check if guest transmits packets when link is up/down<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3347 2010-10-14T11:58:11Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> * Libvirt integration - Will hold on this work item for the next couple of months due to resource constraints.<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Multi host migration - priority increased<br /> <br /> * Set up a vhost job, with the upstream job (mst's trees for kernel and userspace)<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Done ===<br /> <br /> * Resolved autotest bugs that prevented proper kernel build/install functionality for Fedora/newer RHEL systems:<br /> http://patchwork.test.kernel.org/patch/2630/<br /> http://patchwork.test.kernel.org/patch/2631/<br /> http://patchwork.test.kernel.org/patch/2632/<br /> http://patchwork.test.kernel.org/patch/2640/<br /> http://patchwork.test.kernel.org/patch/2634/<br /> http://patchwork.test.kernel.org/patch/2635/<br /> <br /> Now the host kernel install control file patch is fully functional:<br /> http://patchwork.test.kernel.org/patch/2641/<br /> <br /> * Made cache=off the default for KVM autotest tests:<br /> http://patchwork.test.kernel.org/patch/2642/<br /> * Removed boot=on when booting virtio images:<br /> http://patchwork.test.kernel.org/patch/2643/<br /> <br /> * Applied network patchset (Yay!)<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> * Post-commit review of new network patches (committed on Oct 7 2010)<br /> * Use more exceptions in utility functions:<br /> ** Add convenience functions to kvm_subprocess.py to make running shell commands shorter<br /> ** Add convenience functions to the VM class<br /> <br /> == akong ==<br /> <br /> * Report problem with multiple NICs and MAC address tracking<br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge<br /> * Bonding test<br /> * Add a new test: check if guest transmits packets when link is up/down<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3346 2010-10-14T11:53:33Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> * Libvirt integration - Will hold on this work item for the next couple of months due to resource constraints.<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Multi host migration<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Done ===<br /> <br /> * Resolved autotest bugs that prevented proper kernel build/install functionality for Fedora/newer RHEL systems:<br /> http://patchwork.test.kernel.org/patch/2630/<br /> http://patchwork.test.kernel.org/patch/2631/<br /> http://patchwork.test.kernel.org/patch/2632/<br /> http://patchwork.test.kernel.org/patch/2640/<br /> http://patchwork.test.kernel.org/patch/2634/<br /> http://patchwork.test.kernel.org/patch/2635/<br /> <br /> Now the host kernel install control file patch is fully functional:<br /> http://patchwork.test.kernel.org/patch/2641/<br /> <br /> * Made cache=off the default for KVM autotest tests:<br /> http://patchwork.test.kernel.org/patch/2642/<br /> * Removed boot=on when booting virtio images:<br /> http://patchwork.test.kernel.org/patch/2643/<br /> <br /> * Applied network patchset (Yay!)<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> * Post-commit review of new network patches (committed on Oct 7 2010)<br /> * Use more exceptions in utility functions:<br /> ** Add convenience functions to kvm_subprocess.py to make running shell commands shorter<br /> ** Add convenience functions to the VM class<br /> <br /> == akong ==<br /> <br /> * Report problem with multiple NICs and MAC address tracking<br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge<br /> * Bonding test<br /> * Add a new test: check if guest transmits packets when link is up/down<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3345 2010-10-14T04:33:05Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> * Libvirt integration - Will hold on this work item for the next couple of months due to resource constraints.<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Multi host migration<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Done ===<br /> <br /> * Resolved autotest bugs that prevented proper kernel build/install functionality for Fedora/newer RHEL systems:<br /> http://patchwork.test.kernel.org/patch/2630/<br /> http://patchwork.test.kernel.org/patch/2631/<br /> http://patchwork.test.kernel.org/patch/2632/<br /> http://patchwork.test.kernel.org/patch/2640/<br /> http://patchwork.test.kernel.org/patch/2634/<br /> http://patchwork.test.kernel.org/patch/2635/<br /> <br /> Now the host kernel install control file patch is fully functional:<br /> http://patchwork.test.kernel.org/patch/2641/<br /> <br /> * Made cache=off the default for KVM autotest tests:<br /> http://patchwork.test.kernel.org/patch/2642/<br /> * Removed boot=on when booting virtio images:<br /> http://patchwork.test.kernel.org/patch/2643/<br /> <br /> * Applied network patchset (Yay!)<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> * Post-commit review of new network patches (committed on Oct 7 2010)<br /> * Use more exceptions in utility functions:<br /> ** Add convenience functions to kvm_subprocess.py to make running shell commands shorter<br /> ** Add convenience functions to the VM class<br /> <br /> == akong ==<br /> <br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge<br /> * Bonding test<br /> * Add a new test: check if guest transmits packets when link is up/down<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3344 2010-10-14T03:55:10Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> * Libvirt integration - Will hold on this work item for the next couple of months due to resource constraints.<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Make cache=off the default for guest images<br /> * Multi host migration<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Done ===<br /> <br /> * Include support for host kernel install on the KVM default control file - Using autotest standard API to get it done<br /> ** Resolved a bunch of bugs to reach to full functionality<br /> * Apply network patchset (Yay!)<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> * Post-commit review of new network patches (committed on Oct 7 2010)<br /> * Use more exceptions in utility functions:<br /> ** Add convenience functions to kvm_subprocess.py to make running shell commands shorter<br /> ** Add convenience functions to the VM class<br /> <br /> == akong ==<br /> <br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge<br /> * Bonding test<br /> * Add a new test: check if guest transmits packets when link is up/down<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3343 2010-10-14T03:54:28Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> * Libvirt integration - Will hold on this work item for the next couple of months due to resource constraints.<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Make cache=off the default for guest images<br /> * Multi host migration<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Done ===<br /> <br /> * Include support for host kernel install on the KVM default control file - Using autotest standard API to get it done<br /> ** Resolved a bunch of bugs to reach to full functionality<br /> * Apply network patchset (Yay!)<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> * Post-commit review of new network patches (committed on Oct 7 2010)<br /> * Use more exceptions in utility functions:<br /> ** Add convenience functions to kvm_subprocess.py to make running shell commands shorter<br /> ** Add convenience functions to the VM class<br /> <br /> === Very old, possibly no longer relevant ===<br /> <br /> * Add docstrings to all functions that still lack them, including the ones in stepmaker.py, stepeditor.py and kvm_tests.py<br /> * Rename all Windows ISOs currently used to their official MSDN names<br /> * Fill the sections &quot;Working with step files&quot; and &quot;Step file creation tips&quot; in the wiki<br /> <br /> == akong ==<br /> <br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge<br /> * Bonding test<br /> * Add a new test: check if guest transmits packets when link is up/down<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3342 2010-10-14T03:43:39Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == raw idea ==<br /> <br /> * more vhost_net test<br /> * Add Multi-guests transfer test<br /> * transfer test between guest(s) and extra boxs<br /> * add more whql tests (virtio_net,virtio_blk)<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> * Libvirt integration - Will hold on this work item for the next couple of months due to resource constraints.<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Make cache=off the default for guest images<br /> * Multi host migration<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Done ===<br /> <br /> * Include support for host kernel install on the KVM default control file - Using autotest standard API to get it done<br /> ** Resolved a bunch of bugs to reach to full functionality<br /> * Apply network patchset (Yay!)<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> ** This may just be a matter of moving useful code from tests to kvm_test_utils.py to make it reusable<br /> * Post-commit review of new network patches (committed on Oct 7 2010)<br /> * Use more exceptions in utility functions:<br /> ** Add convenience functions to kvm_subprocess.py to make running shell commands shorter<br /> ** Add convenience functions to the VM class<br /> <br /> === Very old, possibly no longer relevant ===<br /> <br /> * Add docstrings to all functions that still lack them, including the ones in stepmaker.py, stepeditor.py and kvm_tests.py<br /> * Rename all Windows ISOs currently used to their official MSDN names<br /> * Fill the sections &quot;Working with step files&quot; and &quot;Step file creation tips&quot; in the wiki<br /> <br /> == akong ==<br /> <br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge<br /> * Bonding test<br /> * Add a new test: check if guest transmits packets when link is up/down<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> (yolkfull)<br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3314 2010-10-07T12:18:48Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> * Libvirt integration - Will hold on this work item for the next couple of months due to resource constraints.<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Make cache=off the default for guest images<br /> * Multi host migration<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Partially done/ Blocked ===<br /> <br /> * Include support for host kernel install on the KVM default control file - Using autotest standard API to get it done<br /> ** needs to resolve bugs in boottool to full functionality<br /> <br /> === Done ===<br /> <br /> * Apply network patchset (Yay!)<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> * Add a way to shutdown the VMs when the whole job completes (compared to doing nothing or shutting-down after every test)<br /> * Add docstrings to all functions that still lack them, including the ones in stepmaker.py, stepeditor.py and kvm_tests.py<br /> * Rename all Windows ISOs currently used to their official MSDN names<br /> * Fill the sections &quot;Working with step files&quot; and &quot;Step file creation tips&quot; in the wiki<br /> <br /> == akong ==<br /> <br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge status: have implemented in internal tree<br /> * Bonding test<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Use private bridge and dnsmasq to do the unattended installation<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3313 2010-10-07T12:18:35Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> * Libvirt integration - Will hold on this work item for the next couple of months due to resource constraint<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Make cache=off the default for guest images<br /> * Multi host migration<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Partially done/ Blocked ===<br /> <br /> * Include support for host kernel install on the KVM default control file - Using autotest standard API to get it done<br /> ** needs to resolve bugs in boottool to full functionality<br /> <br /> === Done ===<br /> <br /> * Apply network patchset (Yay!)<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> * Add a way to shutdown the VMs when the whole job completes (compared to doing nothing or shutting-down after every test)<br /> * Add docstrings to all functions that still lack them, including the ones in stepmaker.py, stepeditor.py and kvm_tests.py<br /> * Rename all Windows ISOs currently used to their official MSDN names<br /> * Fill the sections &quot;Working with step files&quot; and &quot;Step file creation tips&quot; in the wiki<br /> <br /> == akong ==<br /> <br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge status: have implemented in internal tree<br /> * Bonding test<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Use private bridge and dnsmasq to do the unattended installation<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3312 2010-10-07T12:18:23Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == general ==<br /> <br /> '''Note:''' This session is for major, multi-person items. We'll update the status weekly based on the status of the subtasks attributed to persons.<br /> <br /> * Libvirt integration - Will hold on this work item for the next couple of months due to resources<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Make cache=off the default for guest images<br /> * Multi host migration<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Partially done/ Blocked ===<br /> <br /> * Include support for host kernel install on the KVM default control file - Using autotest standard API to get it done<br /> ** needs to resolve bugs in boottool to full functionality<br /> <br /> === Done ===<br /> <br /> * Apply network patchset (Yay!)<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> * Add a way to shutdown the VMs when the whole job completes (compared to doing nothing or shutting-down after every test)<br /> * Add docstrings to all functions that still lack them, including the ones in stepmaker.py, stepeditor.py and kvm_tests.py<br /> * Rename all Windows ISOs currently used to their official MSDN names<br /> * Fill the sections &quot;Working with step files&quot; and &quot;Step file creation tips&quot; in the wiki<br /> <br /> == akong ==<br /> <br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge status: have implemented in internal tree<br /> * Bonding test<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Use private bridge and dnsmasq to do the unattended installation<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr https://linux-kvm.org/index.php?title=KVM-Autotest/TODO&diff=3311 2010-10-07T12:03:56Z

<p>Lmr: </p> <hr /> <div>= KVM-Autotest TODO list =<br /> <br /> == lmr ==<br /> <br /> === TODO ===<br /> <br /> * Make cache=off the default for guest images<br /> * Multi host migration<br /> * Libvirt integration<br /> * Add a BLOCKED status on the autotest database, that tells the user that some feature testing was left out due to a problem on its dependencies.<br /> * Eventually get rid of all dependency on slirp/userspace networking to do any feature testing, since it is not a supported feature. Slipstreaming the kickstart in the install CD is a possibility, and it ends up being simpler than configuring an internal network thing for testing.<br /> * Make it possible to install virtio nw driver under windows xp/2003 guests without having to resort to the msi installer <br /> <br /> === Partially done/ Blocked ===<br /> <br /> * Include support for host kernel install on the KVM default control file - Using autotest standard API to get it done<br /> ** needs to resolve bugs in boottool to full functionality<br /> <br /> === Done ===<br /> <br /> * Apply network patchset (Yay!)<br /> <br /> == mgoldish ==<br /> <br /> * Make WHQL tests run on our internal server<br /> * Enable &quot;guest-load&quot; for a VMs before/while tests are running (e.g. migration of a VM, while a movie is playing on it)<br /> * Add a way to shutdown the VMs when the whole job completes (compared to doing nothing or shutting-down after every test)<br /> * Add docstrings to all functions that still lack them, including the ones in stepmaker.py, stepeditor.py and kvm_tests.py<br /> * Rename all Windows ISOs currently used to their official MSDN names<br /> * Fill the sections &quot;Working with step files&quot; and &quot;Step file creation tips&quot; in the wiki<br /> <br /> == akong ==<br /> <br /> * Using dnsmasq in unattended_install to replace userspace network with private bridge status: have implemented in internal tree<br /> * Bonding test<br /> <br /> == yolkfull ==<br /> <br /> * Run netperf test between two guests<br /> * Migration with/without workload(dbench,lmbench etc)<br /> * Verify SMBIOS/DMI data (UUID, for example)<br /> * Disk serial number (for IDE, SCSI, VirtIO drives)<br /> <br /> == Beijing QE ==<br /> <br /> * pxe boot * net types<br /> * Further migration<br /> * Multiple nics<br /> * multiple disks<br /> * Test block device cancellation path using device mapper to generate errors ( after we had a crash in de_write_dma_cb)<br /> * Extend pci_assignable to support other PCI devices (USB, video cards, TV card etc)<br /> * Different CPU flags support<br /> * Use private bridge and dnsmasq to do the unattended installation<br /> * Passthrough the perf keys of run_autotest to autotest server<br /> * Register the virtual machine into autotest server and run benchmark through autotest server<br /> <br /> <br /> == Bugs ==<br /> <br /> * Please open bugs on the [[http://autotest.kernel.org/newticket autotest defect tracking system]]</div>

Lmr