[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [PATCH v3 4/7] migration/dirtyrate: add per-vcpu option for calc-dir
From: |
Peter Xu |
Subject: |
Re: [PATCH v3 4/7] migration/dirtyrate: add per-vcpu option for calc-dirty-rate |
Date: |
Mon, 7 Jun 2021 13:13:24 -0400 |
On Tue, Jun 08, 2021 at 12:16:03AM +0800, Hyman Huang wrote:
[...]
> > Actually I'm also wondering whether dirty log could be anything useful here
> > in
> > the future as a 3rd mode (then the "*mode" idea should be more useful if
> > so),
> > basically for old kernels where dirty ring is not there, we can timely call
> > memory_global_dirty_log_sync() and calculate dirty pages there just like
> > what
> > we do with dirty rings (without calling migration_bitmap_sync(), so we don't
> > need to deliver dirty bits from kvmslots to ramblocks, just pick them up
> > from
> > kvm and do the accounting for pure dirty rate measurement). That's a wild
> > idea
> > though, so just raise it up. Would that be anything useful?
> uh, actually this idea is what i'm working on to stat the memory heat(trying
> to reduce the transferred memory in migration), theoretically it can be used
> for dirty rate measurement also, maybe i could try it after this patchset
> work having done.
Yes, definitely no need to do that in this series; it just proves that "*mode"
parameter is better in case a 3rd one could come. Thanks,
--
Peter Xu
[PATCH v3 6/7] memory: make global_dirty_log a bitmask, huangy81, 2021/06/06
[PATCH v3 4/7] migration/dirtyrate: add per-vcpu option for calc-dirty-rate, huangy81, 2021/06/06
[PATCH v3 1/7] migration/dirtyrate: make sample page count configurable, huangy81, 2021/06/06
[PATCH v3 2/7] hmp: Add "calc_dirty_rate" and "info dirty_rate" cmds, huangy81, 2021/06/06
[PATCH v3 5/7] migration/dirtyrate: adjust struct DirtyRateStat, huangy81, 2021/06/06
Re: [PATCH v3 0/7] support dirtyrate at the granualrity of vcpu, Hyman, 2021/06/06
Re: [PATCH v3 0/7] support dirtyrate at the granualrity of vcpu, no-reply, 2021/06/09