KAMEZAWA Hiroyuki <kamezawa.hiroyu@xxxxxxxxxxxxxx> wrote:
> On Fri, 24 Jul 2009 14:44:16 +0900 (JST)
> Ryo Tsuruta <ryov@xxxxxxxxxxxxx> wrote:
> good solution to resolve such problem.
> > > My point is "don't allow anyone to use bandwidth of others."
> > > Considering job isolation, a thread who requests swap-out should be charg=
> > > ed
> > > against bandwidth.
> > From another perspective, the swap-out is caused since the buggy
> > process uses a large amount of memory, so it can be considered as
> > the bandwidth of logging process is used due to the buggy process.
> > Please consider the following case. If a thread who requests swap-out
> > is charged, the thread is charged other threads' I/O.
> > (1) -------- (2)
> > Process A | | Process B
> > mmaps a large area in --> | memory | <-- tries to allocate a page.
> > the memory and writes | |
> > data to there. -------- (3)
> > | To get a free page,
> > | the data written by Proc.A
> > | is written out to the disk.
> > V The I/O is done by using
> > --------- Proc.B's bandwidth.
> > | disk |
> > ---------
> > Thus I think that page owners should be charged against bandwidth.
> Ok, no good way. yours is wrong, mine is wrong, too.
> plz find 3rd way, reasonable.
> Below is brief thinking.
> "Why process A should be charged to I/O when it just maps anon memory ?"
> I can't answer this.
> Even in yorr case, Process B requests memory and get penalty. It's
> very natural, I think.
> In usual case,
> - if process A maps ANON, there will be no I/O.
> - if process A maps FILE, it will be charged to process A.
> ok ?
> Under memory pressure,
> - if process A maps ANON, swap I/O should be charged to process B.
> - if process A maps FILE, I/O should be charged to process A.
I think that even process A maps ANON, it should be charged to process A
because the memory pressure is caused by process A. It seems natual
for me that a process which consumes more resources is more likely to
> Anyway, there will be ineraction with dirty_ratio of memcg (not implemeted
> and _Owner should be charged_ issue will be handled in this dirty_ratio layer.
> More consideration is necessary, I think.
I'll keep thinking how it should be done.
Xen-devel mailing list