KVM - Recent changes [en]

Projects/auto-ballooning

LuizCapitulino — Mon, 17 Jun 2013 21:09:58 GMT

Revision as of 21:09, 17 June 2013		Current revision
Line 5:		Line 5:
	When a Linux host is running out of memory, the kernel will take action to ''reclaim'' memory. This action may be detrimental to KVM guests performace (eg. swapping) or even extreme to the point where the kernel may kill a VM or an important virt stack component.		When a Linux host is running out of memory, the kernel will take action to ''reclaim'' memory. This action may be detrimental to KVM guests performace (eg. swapping) or even extreme to the point where the kernel may kill a VM or an important virt stack component.

-	To help avoiding this situation, a KVM guest could automatically return memory to the host when the host is facing memory pressure (automatic inflate). By doing so the guest may also get into memory pressure so we also need a way to allow the guest to automatically get memory back (automatic deflate).	+	To help avoiding this situation, a KVM guest could automatically return memory to the host when the host is facing memory pressure. By doing so the guest may also get into memory pressure so we also need a way to allow the guest to automatically get memory back.

	That's what the automatic ballooning project is about. The next section will dive into ways of implementing it. The following section will provide instructions on how to try current patches.		That's what the automatic ballooning project is about. The next section will dive into ways of implementing it. The following section will provide instructions on how to try current patches.
Line 20:		Line 20:
	=== Automatic Deflate ===		=== Automatic Deflate ===

-	Automatic deflate is performed by the virtio-balloon driver in the guest. There's also two ways of implementing this:	+	Automatic deflate is performed by the virtio-balloon driver within the guest. There's also two ways of implementing this:

	# The virtio-balloon driver registers a callback with the shrinker API. That callback is called when the guest kernel is facing memory pressure. The number of pages to be returned to the kernel is passed to the callback, so that the callback can deflate the balloon by that amount. This is what the current patchset does		# The virtio-balloon driver registers a callback with the shrinker API. That callback is called when the guest kernel is facing memory pressure. The number of pages to be returned to the kernel is passed to the callback, so that the callback can deflate the balloon by that amount. This is what the current patchset does
Line 26:		Line 26:

	== Testing ==		== Testing ==
		+
		+	You have to do three things to play with automatic-ballooning:
		+
		+	# Install kernel 3.10-rc1 or higher in the host. Make sure to enable CONFIG_CGROUPS and CONFIG_MEMCG
		+	# Clone QEMU from ''git://repo.or.cz/qemu/qmp-unstable.git balloon/auto-ballooning/rfc.v2'' (or grab the patch [http://repo.or.cz/w/qemu/qmp-unstable.git/commit/59fec94e6396d6c32cdce47777440c3a988be63d here])
		+	# Install the following kernel on your guest ''git://repo.or.cz/linux-2.6/luiz-linux-2.6.git virtio-balloon/auto-deflate/rfc'' (or grab the patches from the [http://repo.or.cz/w/linux-2.6/luiz-linux-2.6.git/shortlog/refs/heads/virtio-balloon/auto-deflate/rfc web interface])
		+
		+	After setting up the above, do the following to experiment with automatic deflate (which is easy to reproduce):
		+
		+	# Pass ''-balloon virtio,auto-balloon=true'' when starting QEMU
		+	# Wait for the guest to boot, then generate some memory pressure in the guest (say a kernel build with lots of jobs)
		+	# Switch to QEMU's monitor and decrease guest memory (say from 1G to 200MB)
		+	# Watch the guest increase its memory by repeatedly running "info balloon" on QEMU's monitor
		+
		+	To see automatic inflate and automatic deflate in action, you can run several VMs in parallel doing some heavy memory workload. Make sure to over commit host memory. Say your host has 4GB, run 6 VMs with 1GB each.
		+
		+	Another idea is to run one VM, but generate pressure in the host some way like a big kernel build or an application that allocates lots of memory.

OVMF

Lersek — Sat, 15 Jun 2013 13:42:26 GMT

update win2k8r2sp1 guest status

(Difference between revisions)

Projects/auto-ballooning

LuizCapitulino — Thu, 13 Jun 2013 21:11:26 GMT

Revision as of 21:11, 13 June 2013		Current revision
Line 1:		Line 1:
-	= Auto-ballooning =	+	= Automatic Ballooning =

-	When a Linux host is running out of memory, the kernel will take action to ''reclaim'' memory. This action may be detrimental to kvm guests performace (eg. swapping) or even extreme to the point where the kernel kills user-space processes (which could be a kvm guest itself or a virt stack component).	+	== Introduction ==

-	To help avoiding this situation, a kvm guest could automatically inflate its balloon (thus returning memory to the host) when the host is facing memory pressure. Likewise, a kvm guest could automatically deflate its balloon when the guest itself is facing memory pressure (maybe due to automatic inflate).	+	When a Linux host is running out of memory, the kernel will take action to ''reclaim'' memory. This action may be detrimental to KVM guests performace (eg. swapping) or even extreme to the point where the kernel may kill a VM or an important virt stack component.

-	The goal of this page is to collect ideas on how to implement auto-ballooning.	+	To help avoiding this situation, a KVM guest could automatically return memory to the host when the host is facing memory pressure (automatic inflate). By doing so the guest may also get into memory pressure so we also need a way to allow the guest to automatically get memory back (automatic deflate).
-	Currently, there is only one proposal, which is the "automatic balance" idea,	+
-	mostly designed by Rik van Riel.	+

-	== Automatic balance ==	+	That's what the automatic ballooning project is about. The next section will dive into ways of implementing it. The following section will provide instructions on how to try current patches.

-	In automatic balance, the hypervisor (eg. QEMU) must have a way to detect that the host is under memory pressure. When this happens, QEMU could automatically inflate the guest's balloon to some value (say 10MB). When it's the guest that's facing memory pressure, its balloon driver could automatically deflate the memory balloon.	+	== Design ==

-	It's expected that inflate and deflate will balance each other over time.	+	=== Automatic Inflate ===

-	Next sections describe the implementation details of this idea.	+	Automatic inflate is performed by QEMU, that is, the KVM host. There are two ways I'm playing with automatic inflate, in both of them QEMU registers for [http://lwn.net/Articles/544652/ memory pressure] events to learn when the host is under memory pressure:

-	=== auto-inflate ===	+	# QEMU registers for the VMPRESSURE_MEDIUM event and inflates the balloon by an user specified value when the event is received. That value value could default to 16MB or 32MB. This is what the current patchset does
		+	# QEMU registers for the three memory pressure events and inflates the balloon accordingly. Say, it could inflate the balloon by 1MB on VMPRESSURE_LOW; 16MB on VMPRESSURE_MEDIUM and 128MB on VMPRESSURE_CRITICAL. This is not done yet because there's a small limitation in the memory pressure code which doesn't allow this. I'll fix it

-	There are two components needed to make auto-inflate work.	+	=== Automatic Deflate ===

-	* A (host) kernel API to let user-space processes know when the kernel is facing memory pressure	+	Automatic deflate is performed by the virtio-balloon driver in the guest. There's also two ways of implementing this:

-	* QEMU changes to use the host kernel API, and to inflate the guest's balloon when the host kernel is facing memory pressure	+	# The virtio-balloon driver registers a callback with the shrinker API. That callback is called when the guest kernel is facing memory pressure. The number of pages to be returned to the kernel is passed to the callback, so that the callback can deflate the balloon by that amount. This is what the current patchset does
		+	# The virtio-balloon driver registers for an in-kernel memory pressure event (not upstream yet) and deflates the balloon by some fixed amount (maybe the same amount used in automatic inflate)

-	=== auto-deflate ===	+	== Testing ==
-		+
-	Only one component is needed for auto-deflate	+
-		+
-	* The guest virtio-balloon driver must be changed to automatically deflate the balloon when the guest is facing memory pressure	+
-		+
-	=== status ===	+
-		+
-	'''FIXME''': add tree.	+
-		+
-	=== testing ===	+
-		+
-	'''FIXME''': describe.	+

Multiqueue

Kongove — Thu, 13 Jun 2013 05:44:16 GMT

Revision as of 05:44, 13 June 2013		Current revision
Line 214:		Line 214:

	* ...		* ...
		+
		+	== Enable MQ feature ==
		+	* create tap device with multiple queues, please reference
		+	Documentation/networking/tuntap.txt:(3.3 Multiqueue tuntap interface)
		+	* enable mq in qemu cmdline: -device virtio-net-pci,mq=on,...
		+	* enable mq in guest by 'ethtool -L eth0 combined $queue_num'

	== Test ==		== Test ==

NetworkingTodo

Mst — Mon, 10 Jun 2013 07:06:47 GMT

Revision as of 07:06, 10 June 2013		Current revision
Line 105:		Line 105:

	=== projects that are not started yet - no owner ===		=== projects that are not started yet - no owner ===
		+
		+	* Enable GRO for packets coming to bridge from a tap interface
		+
		+	* Better support for windows LRO
		+	Extend virtio-header with statistics for GRO packets:
		+	number of packets coalesced and number of duplicate ACKs coalesced

	* netdev polling for virtio.		* netdev polling for virtio.
Line 195:		Line 201:
	Reduce the number of interrupt		Reduce the number of interrupt

		+	* bridging on top of macvlan
		+	add code to forward LRO status from macvlan (not macvtap)
		+	back to the lowerdev, so that setting up forwarding
		+	from macvlan disables LRO on the lowerdev

		+	* preserve packets exactly with LRO
		+	LRO is not normally compatible with forwarding.
		+	virtio we are getting packets from a linux host,
		+	so we could thinkably preserve packets exactly
		+	even with LRO. I am guessing other hardware could be
		+	doing this as well.

	=== testing projects ===		=== testing projects ===

GuestProgrammableMacVlanFiltering

Kongove — Wed, 05 Jun 2013 11:59:12 GMT

Revision as of 11:59, 5 June 2013		Current revision
Line 8:		Line 8:
	* set promiscuous mode on guest virtio-net device		* set promiscuous mode on guest virtio-net device
	* all this controllable by host admin		* all this controllable by host admin
		+
		+
		+
		+

	TODO:		TODO:
Line 17:		Line 21:

	* Control should be done via qemu/virtio features. Need a way to disable access that qemu can't override unless it has net admin capability.		* Control should be done via qemu/virtio features. Need a way to disable access that qemu can't override unless it has net admin capability.
		+
		+
		+	QEMU:
		+
		+	* Amos Kong works on QEMU side [http://lists.nongnu.org/archive/html/qemu-devel/2013-06/msg00658.html] to add event notification when guest change rx-filter config (main-mac, rx-mode, mac-table, vlan-table). Libvirt will query the rx-filter config from monitor (query-rx-filter), then sync the change to host device.

HOWTO

Tmartinfr — Fri, 31 May 2013 12:49:32 GMT

Add kvm-simple-init to Scripting & Software section

Revision as of 12:49, 31 May 2013		Current revision
Line 86:		Line 86:
	* [http://pve.proxmox.com/wiki/Bare-metal_ISO_Installer Bare-metal] installer with KVM		* [http://pve.proxmox.com/wiki/Bare-metal_ISO_Installer Bare-metal] installer with KVM
	* [[kvmtools\|Python scripts to manage qemu-kvm guest from cmdline]] - yet another qemu-kvm script		* [[kvmtools\|Python scripts to manage qemu-kvm guest from cmdline]] - yet another qemu-kvm script
		+	* [https://github.com/tmartinfr/kvm-simple-init kvm-simple-init] - Simple init script to manage KVM virtual machines

	=== libvirt ===		=== libvirt ===

NetworkingTodo

Rhod — Thu, 30 May 2013 17:10:16 GMT

Revision as of 17:10, 30 May 2013		Current revision
Line 201:		Line 201:

	* Run weekly test on upstream HEAD covering test matrix with autotest		* Run weekly test on upstream HEAD covering test matrix with autotest
		+	* Measure the effect of each of the above-mentioned optimizations
		+	- Use autotest network performance regression testing (that runs netperf)
		+	- Also test any wild idea that works. Some may be useful.

	=== non-virtio-net devices ===		=== non-virtio-net devices ===

Guest Support Status

Hiekjoha — Fri, 24 May 2013 15:54:51 GMT

added Win98SE-3

Revision as of 15:54, 24 May 2013		Current revision
Line 345:		Line 345:
	\| no way		\| no way
	\| install crashes		\| install crashes
		+	\|-
		+
		+	\| Windows 98SE
		+	\| 32
		+	\| kvm-1.12.5
		+	\| AMD
		+	\| 32, 64
		+	\| works with a trick
		+	\| copy an existing installation with qemu-img \
		+	opts: nic,model=ne2k_pci, soundhw=sb16, vga=cirrus+cirrus-Win96-Driver \
		+	tested by hiekjoha
	\|-		\|-

Line 1,108:		Line 1,119:
	\| Works		\| Works
	\|		\|
		+	\|-
		+
		+	\| Knoppix 6.2
		+	\| 32
		+	\| kvm-1.12.5
		+	\| AMD
		+	\| 32, 64
		+	\| Works
		+	\| tested by hiekjoha
		+	\|-
		+
		+	\| Knoppix 6.7.1
		+	\| 32
		+	\| kvm-1.12.5
		+	\| AMD
		+	\| 32, 64
		+	\| Works
		+	\| tested by hiekjoha
	\|}		\|}

Line 2,317:		Line 2,346:
	\| tested by Anna.L Karl /		\| tested by Anna.L Karl /
	[http://www.vittleende.se/ Tandblekning]		[http://www.vittleende.se/ Tandblekning]
		+	\|-
		+
		+	\| clonezilla-live-1.2.* - iso
		+	\| 32
		+	\| kvm-1.12.5
		+	\| AMD
		+	\| 32, 64
		+	\| Works
		+	\| tested by hiekjoha
		+	\|
	\|}		\|}

NetworkingTodo

Mst — Fri, 24 May 2013 14:02:25 GMT

another project with no owner

Revision as of 14:02, 24 May 2013		Current revision
Line 107:		Line 107:

	* netdev polling for virtio.		* netdev polling for virtio.
		+	There are two kinds of netdev polling:
		+	- netpoll - used for debugging
		+	- proposed low latency net polling
	See http://lkml.indiana.edu/hypermail/linux/kernel/1303.0/00553.html		See http://lkml.indiana.edu/hypermail/linux/kernel/1303.0/00553.html

Line 133:		Line 136:
	* more GSO type support:		* more GSO type support:
	Kernel not support more type of GSO: FCOE, GRE, UDP_TUNNEL		Kernel not support more type of GSO: FCOE, GRE, UDP_TUNNEL
		+
		+	* ring aliasing:
		+	using vhost-net as a networking backend with virtio-net in QEMU
		+	being what's guest facing.
		+	This gives you the best of both worlds: QEMU acts as a first
		+	line of defense against a malicious guest while still getting the
		+	performance advantages of vhost-net (zero-copy).
		+	In fact a bit of complexity in vhost was put there in the vague hope to
		+	support something like this: virtio rings are not translated through
		+	regular memory tables, instead, vhost gets a pointer to ring address.
		+	This allows qemu acting as a man in the middle,
		+	verifying the descriptors but not touching the packet data.
		+

	=== vague ideas: path to implementation not clear		=== vague ideas: path to implementation not clear

NetworkingTodo

Jasowang — Fri, 24 May 2013 09:34:53 GMT

Revision as of 09:34, 24 May 2013		Current revision
Line 23:		Line 23:
	Multiqueue causes regression in some workloads, thus		Multiqueue causes regression in some workloads, thus
	it is off by default. Detect and enable/disable		it is off by default. Detect and enable/disable
-	automatically so we can make it on by default	+	automatically so we can make it on by default.
		+	This is because GSO tends to batch less when mq is enabled.
		+	https://patchwork.kernel.org/patch/2235191/
		+	Developer: Jason Wang
		+
		+	* rework on flow caches
		+	Current hlist implementation of flow caches has several limitations:
		+	1) at worst case, linear search will be bad
		+	2) not scale
		+	https://patchwork.kernel.org/patch/2025121/
		+	Developer: Jason Wang
		+
		+	* eliminate the extra copy in virtio-net driver
		+	We need do an extra copy of 128 bytes for every packets.
		+	This could be eliminated for small packets by:
		+	1) use build_skb() and head frag
		+	2) bigger vnet header length ( >= NET_SKB_PAD + NET_IP_ALIGN )
		+	Or use a dedicated queue for small packet receiving ? (reordering)
		+	Developer: Jason Wang
		+
		+	* make pktgen works for virtio-net ( or partially orphan )
		+	virtio-net orphan the skb during tx,
		+	which will makes pktgen wait for ever to the refcnt.
		+	Jason's idea: introduce a flat to tell pktgen not for wait
		+	Discussion here: https://patchwork.kernel.org/patch/1800711/
		+	MST's idea: add a .ndo_tx_polling not only for pktgen
		+	Developer: Jason Wang
		+
		+	* Add HW_VLAN_TX support for tap
		+	Eliminate the extra data moving for tagged packets
		+	Developer: Jason Wang
		+
		+	* Announce self by guest driver
		+	Send gARP by guest driver. Guest part is finished.
		+	Qemu is ongoing.
		+	V7 patches is here:
		+	http://lists.nongnu.org/archive/html/qemu-devel/2013-03/msg01127.html
	Developer: Jason Wang		Developer: Jason Wang

Line 94:		Line 130:
	support "reposting" buffers for mergeable buffers,		support "reposting" buffers for mergeable buffers,
	support pool for indirect buffers		support pool for indirect buffers
		+
		+	* more GSO type support:
		+	Kernel not support more type of GSO: FCOE, GRE, UDP_TUNNEL

	=== vague ideas: path to implementation not clear		=== vague ideas: path to implementation not clear
Line 108:		Line 147:
	core, need to teach it to allocate array of		core, need to teach it to allocate array of
	pointers and not array of queues.		pointers and not array of queues.
		+	Jason has an draft patch to use flex array.
		+	Another thing is to move the flow caches out of tun_struct.
		+	Developer: Jason Wang

	* irq/numa affinity:		* irq/numa affinity:
Line 130:		Line 172:
	* vlan filtering in bridge		* vlan filtering in bridge
	IGMP snooping in bridge should take vlans into account		IGMP snooping in bridge should take vlans into account
		+
		+	* tx coalescing
		+	Delay several packets before kick the device.
		+
		+	* interrupt coalescing
		+	Reduce the number of interrupt
		+

NetworkingTodo

Mst — Thu, 23 May 2013 21:52:46 GMT

Revision as of 21:52, 23 May 2013		Current revision
Line 8:		Line 8:
	* vhost-net scalability tuning: threading for many VMs		* vhost-net scalability tuning: threading for many VMs
	Plan: switch to workqueue shared by many VMs		Plan: switch to workqueue shared by many VMs
-	www.mail-archive.com/kvm@vger.kernel.org/msg69868.html	+	http://www.mail-archive.com/kvm@vger.kernel.org/msg69868.html
-	Developer: Shirley Ma?, MST	+
		+	http://domino.research.ibm.com/library/cyberdig.nsf/1e4115aea78b6e7c85256b360066f0d4/479e3578ed05bfac85257b4200427735!OpenDocument
		+
		+	Developer: Shirley Ma?, MST?
	Testing: netperf guest to guest		Testing: netperf guest to guest

Line 79:		Line 82:

	* RDMA bridging		* RDMA bridging
		+
		+	* DMA emgine (IOAT) use in tun
		+	Old patch here: [PATCH RFC] tun: dma engine support
		+	It does not speed things up. Need to see why and
		+	what can be done.

	* use kvm eventfd support for injecting level interrupts,		* use kvm eventfd support for injecting level interrupts,
	enable vhost by default for level interrupts		enable vhost by default for level interrupts
-
-	* DMA emgine (IOAT) use in tun

	* virtio API extension: improve small packet/large buffer performance:		* virtio API extension: improve small packet/large buffer performance:
Line 98:		Line 104:

	* support more queues		* support more queues
-	We limit TUN to 8 queues	+	We limit TUN to 8 queues, but we really want
		+	1 queue per guest CPU. The limit comes from net
		+	core, need to teach it to allocate array of
		+	pointers and not array of queues.

	* irq/numa affinity:		* irq/numa affinity:

KVM Forum 2012

Knoel — Thu, 23 May 2013 18:49:41 GMT

Revision as of 18:49, 23 May 2013		Current revision
Line 61:		Line 61:
	\|2:30pm \|\| [[Media:2012-forum-glusterfs-stefanha.pdf \| GlusterFS for KVM Users and Developers]] \|\| Stefan Hajnoczi, Red Hat \|\| [[Media:2012-forum-megasas-vfio.pdf \| Megasas on Steroids: Qemu Device-Passthrough With VFIO]] \|\| Hannes Reinecke, SUSE		\|2:30pm \|\| [[Media:2012-forum-glusterfs-stefanha.pdf \| GlusterFS for KVM Users and Developers]] \|\| Stefan Hajnoczi, Red Hat \|\| [[Media:2012-forum-megasas-vfio.pdf \| Megasas on Steroids: Qemu Device-Passthrough With VFIO]] \|\| Hannes Reinecke, SUSE
	\|-		\|-
-	\|3:30pm \|\| [[Media:2012-forum-nakajima apicv.pdf \| Enabling Optimized Interrupt/APIC Virtualization in KVM]] \|\| Jun Nakajima, Intel \|\| QEMU Live Block Operations: Snapshots, Merging, and Mirroring \|\| Jeff Cody, Red Hat	+	\|3:30pm \|\| [[Media:2012-forum-nakajima apicv.pdf \| Enabling Optimized Interrupt/APIC Virtualization in KVM]] \|\| Jun Nakajima, Intel \|\| [[Media:2012-forum-Snapshots-Cody.pdf \| QEMU Live Block Operations: Snapshots, Merging, and Mirroring]] \|\| Jeff Cody, Red Hat
	\|-		\|-
	\|4:00pm \|\| [[Media:2012-forum-device-sharing-Bharat.pdf \| Efficient Sharing of Physical Devices Between KVM Guests and Host]] \|\| Bharat Bhushan, Freescale \|\| [[Media:2012-forum-virtio-blk-performance-improvement.pdf \| Virtio-blk Performance Improvement]] \|\| Asias He, Red Hat		\|4:00pm \|\| [[Media:2012-forum-device-sharing-Bharat.pdf \| Efficient Sharing of Physical Devices Between KVM Guests and Host]] \|\| Bharat Bhushan, Freescale \|\| [[Media:2012-forum-virtio-blk-performance-improvement.pdf \| Virtio-blk Performance Improvement]] \|\| Asias He, Red Hat

Special:Log/upload

Knoel — Thu, 23 May 2013 18:47:46 GMT

uploaded "[[Image:2012-forum-Snapshots-Cody.pdf]]"

Main Page

Knoel — Thu, 23 May 2013 16:36:21 GMT

Revision as of 16:36, 23 May 2013		Current revision
Line 13:		Line 13:
	\|style="width:50%;border:none;"\|		\|style="width:50%;border:none;"\|
	* [[KVM Forum]]		* [[KVM Forum]]
		+	** [[KVM Forum 2012]]
	** [[KVM Forum 2011]]		** [[KVM Forum 2011]]
	** [[KVM Forum 2010]]		** [[KVM Forum 2010]]

NetworkingTodo

Kongove — Thu, 23 May 2013 13:00:19 GMT

Revision as of 13:00, 23 May 2013		Current revision
Line 61:		Line 61:
	* Improve stats, make them more helpful for per analysis		* Improve stats, make them more helpful for per analysis
	Developer: Sriram Narasimhan		Developer: Sriram Narasimhan
		+
		+	* Bug: e1000 & rtl8139: Change macaddr in guest, but not update to qemu (info network)
		+	Developer: Amos Kong

	=== projects that are not started yet - no owner ===		=== projects that are not started yet - no owner ===

NetworkingTodo

Mst — Thu, 23 May 2013 10:43:24 GMT

Revision as of 10:43, 23 May 2013		Current revision
Line 2:		Line 2:
	currently most info is related to virtio-net.		currently most info is related to virtio-net.

-	Stabilization is highest priority currently.	+	TODO: add bugzilla entry links.
-	DOA test matrix (all combinations should work):	+
-	vhost: test both on and off, obviously	+
-	test: hotplug/unplug, vlan/mac filtering, netperf,	+
-	file copy both ways: scp, NFS, NTFS	+
-	guests: linux: release and debug kernels, windows	+
-	conditions: plain run, run while under migration,	+
-	vhost on/off migration	+
-	networking setup: simple, qos with cgroups	+
-	host configuration: host-guest, external-guest	+

-	=== vhost-net driver projects ===	+	=== projects in progress. contributions are still very wellcome!
-	* iovec length limitations	+
-	Developer: Jason Wang <jasowang@redhat.com>	+
-	Testing: guest to host file transfer on windows.	+

-	* mergeable buffers: fix host->guest BW regression	+	* vhost-net scalability tuning: threading for many VMs
-	Testing: netperf host to guest default flags	+	Plan: switch to workqueue shared by many VMs
-		+	www.mail-archive.com/kvm@vger.kernel.org/msg69868.html
-	* scalability tuning: threading for guest to guest	+	Developer: Shirley Ma?, MST
-	Developer: MST	+
	Testing: netperf guest to guest		Testing: netperf guest to guest

-	=== qemu projects ===	+	* multiqueue support in macvtap
-	* fix hotplug issues	+	multiqueue is only supported for tun.
-	Developer: MST	+	Add support for macvtap.
-	https://bugzilla.redhat.com/show_bug.cgi?id=623735	+	Developer: Jason Wang
-		+
-	* migration with multiple macs/vlans	+
-	Developer: Jason Wang	+
-	qemu only sends ping with the first mac/no vlan:	+
-	need to send it for all macs/vlan	+

-	* bugfix: crash with illegal fd= value on command line	+	* enable multiqueue by default
		+	Multiqueue causes regression in some workloads, thus
		+	it is off by default. Detect and enable/disable
		+	automatically so we can make it on by default
	Developer: Jason Wang		Developer: Jason Wang
-	Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=581750

-	=== virtio projects ===	+	* guest programmable mac/vlan filtering with macvtap
-	* suspend/resume support	+	Developer: Dragos Tatulea?, Amos Kong
		+	Status: [[GuestProgrammableMacVlanFiltering]]

-	* API extension: improve small packet/large buffer performance:	+	* bridge without promisc mode in NIC
		+	given hardware support, teach bridge
		+	to program mac/vlan filtering in NIC
		+	Helps performance and security on noisy LANs
		+	http://comments.gmane.org/gmane.linux.network/266546
		+	Developer: Vlad Yasevich
		+
		+	* reduce networking latency:
		+	allow handling short packets from softirq or VCPU context
		+	Plan:
		+	We are going through the scheduler 3 times
		+	(could be up to 5 if softirqd is involved)
		+	Consider RX: host irq -> io thread -> VCPU thread ->
		+	guest irq -> guest thread.
		+	This adds a lot of latency.
		+	We can cut it by some 1.5x if we do a bit of work
		+	either in the VCPU or softirq context.
		+	Testing: netperf TCP RR - should be improved drastically
		+	netperf TCP STREAM guest to host - no regression
		+	Developer: MST
		+
		+	* Flexible buffers: put virtio header inline with packet data
		+	https://patchwork.kernel.org/patch/1540471/
		+	Developer: MST
		+
		+	* device failover to allow migration with assigned devices
		+	https://fedoraproject.org/wiki/Features/Virt_Device_Failover
		+	Developer: Gal Hammer, Cole Robinson, Laine Stump, MST
		+
		+	* Reuse vringh code for better maintainability
		+	Developer: Rusty Russell
		+
		+	* Improve stats, make them more helpful for per analysis
		+	Developer: Sriram Narasimhan
		+
		+	=== projects that are not started yet - no owner ===
		+
		+	* netdev polling for virtio.
		+	See http://lkml.indiana.edu/hypermail/linux/kernel/1303.0/00553.html
		+
		+	* receive side zero copy
		+	The ideal is a NIC with accelerated RFS support,
		+	So we can feed the virtio rx buffers into the correct NIC queue.
		+	Depends on non promisc NIC support in bridge.
		+
		+	* IPoIB infiniband bridging
		+	Plan: implement macvtap for ipoib and virtio-ipoib
		+
		+	* RDMA bridging
		+
		+	* use kvm eventfd support for injecting level interrupts,
		+	enable vhost by default for level interrupts
		+
		+	* DMA emgine (IOAT) use in tun
		+
		+	* virtio API extension: improve small packet/large buffer performance:
	support "reposting" buffers for mergeable buffers,		support "reposting" buffers for mergeable buffers,
	support pool for indirect buffers		support pool for indirect buffers
		+
		+	=== vague ideas: path to implementation not clear
		+
	* ring redesign:		* ring redesign:
	find a way to test raw ring performance		find a way to test raw ring performance
	fix cacheline bounces		fix cacheline bounces
	reduce interrupts		reduce interrupts
-	Developer: MST
-	see patchset: virtio: put last seen used index into ring itself

-	=== projects involing other kernel components and/or networking stack ===
-	* guest programmable mac/vlan filtering with macvtap
-	Developer: Dragos Tatulea
-	Status: [[GuestProgrammableMacVlanFiltering]]

-	* bridge without promisc mode in NIC	+	* support more queues
-	given hardware support, teach bridge	+	We limit TUN to 8 queues
-	to program mac/vlan filtering in NIC	+
		+	* irq/numa affinity:
		+	networking goes much faster with irq pinning:
		+	both with and without numa.
		+	what can be done to make the non-pinned setup go faster?
		+
		+	* reduce conflict with VCPU thread
		+	if VCPU and networking run on same CPU,
		+	they conflict resulting in bad performance.
		+	Fix that, push vhost thread out to another CPU
		+	more aggressively.

	* rx mac filtering in tun		* rx mac filtering in tun
Line 68:		Line 115:
	* vlan filtering in tun		* vlan filtering in tun
	the need for this is still not understood as we have filtering in bridge		the need for this is still not understood as we have filtering in bridge
-	for small # if vlans we can use BPF

	* vlan filtering in bridge		* vlan filtering in bridge
	IGMP snooping in bridge should take vlans into account		IGMP snooping in bridge should take vlans into account

-	* zero copy tx/rx for macvtap
-	Developers: tx zero copy Shirley Ma; rx zero copy Xin Xiaohui
-
-	* multiqueue (involves all of vhost, qemu, virtio, networking stack)
-	Developer: Krishna Jumar
-	Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=632751
-
-	* kvm MSI interrupt injection fast path
-	Developer: MST
-
-	* kvm eventfd support for injecting level interrupts
-
-	* DMA emgine (IOAT) use in tun
-
-	* allow handling short packets from softirq context
-	Testing: netperf TCP STREAM guest to host
-	netperf TCP RR
-
-	* irq affinity:
-	networking goes much faster with irq pinning:
-	both with and without numa.
-	what can be done to make the non-pinned setup go faster?

	=== testing projects ===		=== testing projects ===
-	* Cover test matrix with autotest	+	Keeping networking stable is highest priority.
-	* Test with windows drivers, pass WHQL	+
		+	* Run weekly test on upstream HEAD covering test matrix with autotest

	=== non-virtio-net devices ===		=== non-virtio-net devices ===
	* e1000: stabilize		* e1000: stabilize
-	Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=602205

-	=== bugzilla entries for bugs fixed ===	+	=== test matrix ===
-	* verify these are ok upstream	+
-	https://bugzilla.redhat.com/show_bug.cgi?id=623552	+
-	https://bugzilla.redhat.com/show_bug.cgi?id=632747	+
-	https://bugzilla.redhat.com/show_bug.cgi?id=632745	+

-		+	DOA test matrix (all combinations should work):
-	=== abandoned projects: ===	+	vhost: test both on and off, obviously
-	* Add GSO/checksum offload support to AF_PACKET(raw) sockets.	+	test: hotplug/unplug, vlan/mac filtering, netperf,
-	status: incomplete	+	file copy both ways: scp, NFS, NTFS
-	* guest kernel 2.6.31 seems to work well. Under certain workloads,	+	guests: linux: release and debug kernels, windows
-	virtio performance has regressed with guest kernels 2.6.32 and up	+	conditions: plain run, run while under migration,
-	(but still better than userspace). A patch has been posted:	+	vhost on/off migration
-	http://www.spinics.net/lists/netdev/msg115292.html	+	networking setup: simple, qos with cgroups
-	status: might be fixed, need to test	+	host configuration: host-guest, external-guest

Main Page

Javantea — Tue, 21 May 2013 18:29:53 GMT

Remove spam

Revision as of 18:29, 21 May 2013		Current revision
Line 29:		Line 29:
	\|style="width:50%;border:none;"\|		\|style="width:50%;border:none;"\|
	* [http://www.qemu.org/ QEMU]		* [http://www.qemu.org/ QEMU]
-	* [http://www.diamondlinks.net link building service]
-	* [http://www.vinfotech.com/solutions/social-network-design.htm Social Network Design]
-	* [http://www.vinfotech.com/solutions/startup-web-design.htm Startup Web Design]
-	* [http://www.vinfotech.com/web_2.0/web_20_design.htm Web 2.0 Design]
-	* [http://impact-resumes.net/services/career-change-resume.htm Career Change Resume]
	* [http://wiki.xensource.com/xenwiki/HVM_Compatible_Processors Xen's HVM Compatible Processors List]		* [http://wiki.xensource.com/xenwiki/HVM_Compatible_Processors Xen's HVM Compatible Processors List]
	\|style="width:50%;border:none;"\|		\|style="width:50%;border:none;"\|