* Re: [Qemu-devel] [Bug 1100843] Re: Live Migration Causes Performance Issues [not found] ` <20130926203354.30826.10562.malone@soybean.canonical.com> @ 2013-10-06 13:57 ` Zhang Haoyu 2013-10-07 6:38 ` Peter Lieven 0 siblings, 1 reply; 6+ messages in thread From: Zhang Haoyu @ 2013-10-06 13:57 UTC (permalink / raw) To: Bug 1100843, gleb, eblake, quintela, pbonzini, afaerber, xiaoguangrong, mtosatti, kvm, mst Cc: qemu-devel >>From my testing this has been fixed in the saucy version (1.5.0) of qemu. It is fixed by this patch: >f1c72795af573b24a7da5eb52375c9aba8a37972 > >However later in the history this commit was reverted, and again broke this. The other commit that fixes this is: >211ea74022f51164a7729030b28eec90b6c99a08 > See below post,please. https://lists.gnu.org/archive/html/qemu-devel/2013-08/msg05062.html Thanks, Zhang Haoyu >So 211ea740 needs to be backported to P/Q/R to fix this issue. I have a v1 packages of a precise backport here, I've confirmed performance differences between savevm/loadvm cycles: >http://people.canonical.com/~arges/lp1100843/precise/ > >** No longer affects: linux (Ubuntu) > >** Also affects: qemu-kvm (Ubuntu Precise) > Importance: Undecided > Status: New > >** Also affects: qemu-kvm (Ubuntu Quantal) > Importance: Undecided > Status: New > >** Also affects: qemu-kvm (Ubuntu Raring) > Importance: Undecided > Status: New > >** Also affects: qemu-kvm (Ubuntu Saucy) > Importance: High > Assignee: Chris J Arges (arges) > Status: In Progress > >** Changed in: qemu-kvm (Ubuntu Precise) > Assignee: (unassigned) => Chris J Arges (arges) > >** Changed in: qemu-kvm (Ubuntu Quantal) > Assignee: (unassigned) => Chris J Arges (arges) > >** Changed in: qemu-kvm (Ubuntu Raring) > Assignee: (unassigned) => Chris J Arges (arges) > >** Changed in: qemu-kvm (Ubuntu Precise) > Importance: Undecided => High > >** Changed in: qemu-kvm (Ubuntu Quantal) > Importance: Undecided => High > >** Changed in: qemu-kvm (Ubuntu Raring) > Importance: Undecided => High > >** Changed in: qemu-kvm (Ubuntu Saucy) > Assignee: Chris J Arges (arges) => (unassigned) > >** Changed in: qemu-kvm (Ubuntu Saucy) > Status: In Progress => Fix Released > >** Changed in: qemu-kvm (Ubuntu Raring) > Status: New => Triaged > >** Changed in: qemu-kvm (Ubuntu Quantal) > Status: New => Triaged > >** Changed in: qemu-kvm (Ubuntu Precise) > Status: New => In Progress ^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [Bug 1100843] Re: Live Migration Causes Performance Issues 2013-10-06 13:57 ` [Qemu-devel] [Bug 1100843] Re: Live Migration Causes Performance Issues Zhang Haoyu @ 2013-10-07 6:38 ` Peter Lieven 2013-10-07 9:37 ` [Qemu-devel] " Paolo Bonzini 0 siblings, 1 reply; 6+ messages in thread From: Peter Lieven @ 2013-10-07 6:38 UTC (permalink / raw) To: Zhang Haoyu, Bug 1100843, gleb, eblake, quintela, pbonzini, afaerber, xiaoguangrong, mtosatti, kvm, mst Cc: qemu-devel On 06.10.2013 15:57, Zhang Haoyu wrote: >> >From my testing this has been fixed in the saucy version (1.5.0) of > qemu. It is fixed by this patch: >> f1c72795af573b24a7da5eb52375c9aba8a37972 >> >> However later in the history this commit was reverted, and again broke > this. The other commit that fixes this is: >> 211ea74022f51164a7729030b28eec90b6c99a08 >> > See below post,please. > https://lists.gnu.org/archive/html/qemu-devel/2013-08/msg05062.html I would still like to fix qemu to not load roms etc. if we set up a migration target. In this case we could drop the madvise, skip the checking for zero pages and also could avoid sending zero pages at all. It would be the cleanest solution. Peter ^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [Qemu-devel] [Bug 1100843] Re: Live Migration Causes Performance Issues 2013-10-07 6:38 ` Peter Lieven @ 2013-10-07 9:37 ` Paolo Bonzini 2013-10-07 9:49 ` Peter Lieven 0 siblings, 1 reply; 6+ messages in thread From: Paolo Bonzini @ 2013-10-07 9:37 UTC (permalink / raw) To: Peter Lieven Cc: Zhang Haoyu, Bug 1100843, gleb, eblake, quintela, afaerber, xiaoguangrong, mtosatti, kvm, mst, qemu-devel Il 07/10/2013 08:38, Peter Lieven ha scritto: > On 06.10.2013 15:57, Zhang Haoyu wrote: >>> >From my testing this has been fixed in the saucy version (1.5.0) of >> qemu. It is fixed by this patch: >>> f1c72795af573b24a7da5eb52375c9aba8a37972 >>> >>> However later in the history this commit was reverted, and again broke >> this. The other commit that fixes this is: >>> 211ea74022f51164a7729030b28eec90b6c99a08 >>> >> See below post,please. >> https://lists.gnu.org/archive/html/qemu-devel/2013-08/msg05062.html > > I would still like to fix qemu to not load roms etc. if we set up a > migration target. In this case > we could drop the madvise, skip the checking for zero pages and also > could avoid sending > zero pages at all. It would be the cleanest solution. It's in general not easy to do this if you take non-x86 targets into account. Paolo ^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [Qemu-devel] [Bug 1100843] Re: Live Migration Causes Performance Issues 2013-10-07 9:37 ` [Qemu-devel] " Paolo Bonzini @ 2013-10-07 9:49 ` Peter Lieven 2013-10-07 9:55 ` Paolo Bonzini 0 siblings, 1 reply; 6+ messages in thread From: Peter Lieven @ 2013-10-07 9:49 UTC (permalink / raw) To: Paolo Bonzini Cc: Zhang Haoyu, Bug 1100843, gleb, eblake, quintela, afaerber, xiaoguangrong, mtosatti, kvm, mst, qemu-devel On 07.10.2013 11:37, Paolo Bonzini wrote: > Il 07/10/2013 08:38, Peter Lieven ha scritto: >> On 06.10.2013 15:57, Zhang Haoyu wrote: >>>> >From my testing this has been fixed in the saucy version (1.5.0) of >>> qemu. It is fixed by this patch: >>>> f1c72795af573b24a7da5eb52375c9aba8a37972 >>>> >>>> However later in the history this commit was reverted, and again broke >>> this. The other commit that fixes this is: >>>> 211ea74022f51164a7729030b28eec90b6c99a08 >>>> >>> See below post,please. >>> https://lists.gnu.org/archive/html/qemu-devel/2013-08/msg05062.html >> I would still like to fix qemu to not load roms etc. if we set up a >> migration target. In this case >> we could drop the madvise, skip the checking for zero pages and also >> could avoid sending >> zero pages at all. It would be the cleanest solution. > It's in general not easy to do this if you take non-x86 targets into > account. What about the dirty way to zero out all non zero pages at the beginning of ram_load? Peter ^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [Qemu-devel] [Bug 1100843] Re: Live Migration Causes Performance Issues 2013-10-07 9:49 ` Peter Lieven @ 2013-10-07 9:55 ` Paolo Bonzini 2013-10-10 8:17 ` Peter Lieven 0 siblings, 1 reply; 6+ messages in thread From: Paolo Bonzini @ 2013-10-07 9:55 UTC (permalink / raw) To: Peter Lieven Cc: Zhang Haoyu, Bug 1100843, gleb, eblake, quintela, afaerber, xiaoguangrong, mtosatti, kvm, mst, qemu-devel Il 07/10/2013 11:49, Peter Lieven ha scritto: >> It's in general not easy to do this if you take non-x86 targets into >> account. > What about the dirty way to zero out all non zero pages at the beginning of > ram_load? I'm not sure I follow? Paolo ^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [Bug 1100843] Re: Live Migration Causes Performance Issues 2013-10-07 9:55 ` Paolo Bonzini @ 2013-10-10 8:17 ` Peter Lieven 0 siblings, 0 replies; 6+ messages in thread From: Peter Lieven @ 2013-10-10 8:17 UTC (permalink / raw) To: Paolo Bonzini Cc: kvm, gleb, quintela, mtosatti, qemu-devel, Zhang Haoyu, xiaoguangrong, Bug 1100843, mst, afaerber On 07.10.2013 11:55, Paolo Bonzini wrote: > Il 07/10/2013 11:49, Peter Lieven ha scritto: >>> It's in general not easy to do this if you take non-x86 targets into >>> account. >> What about the dirty way to zero out all non zero pages at the beginning of >> ram_load? > I'm not sure I follow? sth like this for each ram block at the beginning of ram_load. + base = memory_region_get_ram_ptr(block->mr); + for (offset = 0; offset < block->length; + offset += TARGET_PAGE_SIZE) { + if (!is_zero_page(base + offset)) { + memset(base + offset, 0x00, TARGET_PAGE_SIZE); + } + } + Then add a capability "skip_zero_pages" which does not sent them on the source and enables this zeroing. it would also be possible to skip the zero check for each incoming compressed pages. Peter ^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2013-10-10 8:17 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
[not found] <20130117163740.7157.55600.malonedeb@gac.canonical.com>
[not found] ` <20130926203354.30826.10562.malone@soybean.canonical.com>
2013-10-06 13:57 ` [Qemu-devel] [Bug 1100843] Re: Live Migration Causes Performance Issues Zhang Haoyu
2013-10-07 6:38 ` Peter Lieven
2013-10-07 9:37 ` [Qemu-devel] " Paolo Bonzini
2013-10-07 9:49 ` Peter Lieven
2013-10-07 9:55 ` Paolo Bonzini
2013-10-10 8:17 ` Peter Lieven
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for NNTP newsgroup(s).