From: Tyler Sanderson <tysand@google.com>
To: David Hildenbrand <david@redhat.com>
Cc: virtualization@lists.linux-foundation.org,
Wei Wang <wei.w.wang@intel.com>,
David Rientjes <rientjes@google.com>,
"linux-mm@kvack.org" <linux-mm@kvack.org>,
Michal Hocko <mhocko@kernel.org>,
"Michael S. Tsirkin" <mst@redhat.com>
Subject: Re: Balloon pressuring page cache
Date: Thu, 30 Jan 2020 14:46:21 -0800 [thread overview]
Message-ID: <CAJuQAmq3xPb5k=egCHwwYD=so8hwOCf6d42mUvVXOh2iLJ64dg@mail.gmail.com> (raw)
In-Reply-To: <75d4594f-0864-5172-a0f8-f97affedb366@redhat.com>
[-- Attachment #1: Type: text/plain, Size: 3072 bytes --]
On Thu, Jan 30, 2020 at 7:03 AM David Hildenbrand <david@redhat.com> wrote:
> On 29.01.20 20:11, Tyler Sanderson wrote:
> >
> >
> > On Wed, Jan 29, 2020 at 2:31 AM David Hildenbrand <david@redhat.com
> > <mailto:david@redhat.com>> wrote:
> >
> > On 29.01.20 01:22, Tyler Sanderson via Virtualization wrote:
> > > A primary advantage of virtio balloon over other memory reclaim
> > > mechanisms is that it can pressure the guest's page cache into
> > shrinking.
> > >
> > > However, since the balloon driver changed to using the shrinker API
> > >
> > <
> https://github.com/torvalds/linux/commit/71994620bb25a8b109388fefa9e99a28e355255a#diff-fd202acf694d9eba19c8c64da3e480c9
> > this
> > > use case has become a bit more tricky. I'm wondering what the
> intended
> > > device implementation is.
> > >
> > > When inflating the balloon against page cache (i.e. no free memory
> > > remains) vmscan.c will both shrink page cache, but also invoke the
> > > shrinkers -- including the balloon's shrinker. So the balloon
> driver
> > > allocates memory which requires reclaim, vmscan gets this memory by
> > > shrinking the balloon, and then the driver adds the memory back to
> the
> > > balloon. Basically a busy no-op.
> > >
> > > If file IO is ongoing during this balloon inflation then the page
> > cache
> > > could be growing which further puts "back pressure" on the balloon
> > > trying to inflate. In testing I've seen periods of > 45 seconds
> where
> > > balloon inflation makes no net forward progress.
> > >
> > > This wasn't a problem before the change to the shrinker API since
> > forced
> > > balloon deflation only occurred via the OOM notifier callback
> > which was
> > > invoked only after the page cache had depleted.
> > >
> > > Is this new busy behavior working as intended?
> >
> > Please note that the shrinker will only be registered in case we have
> > VIRTIO_BALLOON_F_DEFLATE_ON_OOM - (which is AFAIK very rare) - to
> > implement automatic balloon deflation when the guest is under memory
> > pressure.
> >
> >
> > Are you actually experiencing issues with that or did you just
> stumble
> > over the code?
> >
> >
> > We have a use case that is encountering this (and that registers
> > DEFLATE_ON_OOM). We can work around this, but it does seem inefficient.
> > I understand there were good reasons for moving away from the OOM
> > notifier callback, but I'm wondering if the balloon driver could specify
> > a "nice" level to the shrinker API that would cause it to be reclaimed
> > from only as a last resort?
> >
>
> Cc-ing linux-mm, Michal and Michael.
>
> Just wondering, how does your workaround look like?
>
The work around is to monitor the memory statistics reported on the stats
queue. Keep inflating (inefficiently) -- despite the activity on the
deflate queue -- until memory statistics indicate the guest is actually low
on available memory.
--
> Thanks,
>
> David / dhildenb
>
>
[-- Attachment #2: Type: text/html, Size: 4413 bytes --]
prev parent reply other threads:[~2020-01-30 22:46 UTC|newest]
Thread overview: 59+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <CAJuQAmpDUyve2S+oxp9tLUhuRcnddXnNztC5PmYOOCpY6c68xg@mail.gmail.com>
[not found] ` <91270a68-ff48-88b0-219c-69801f0c252f@redhat.com>
[not found] ` <CAJuQAmoaK0Swytu2Os_SQRfG5_LqiCPaDa9yatatm9MtfncNTQ@mail.gmail.com>
2020-01-30 15:02 ` Balloon pressuring page cache David Hildenbrand
2020-01-30 15:20 ` Michael S. Tsirkin
2020-01-30 15:23 ` David Hildenbrand
2020-01-30 15:31 ` Wang, Wei W
2020-01-30 19:59 ` Tyler Sanderson
2020-02-03 13:11 ` Michael S. Tsirkin
2020-02-03 16:18 ` Alexander Duyck
2020-02-03 16:34 ` David Hildenbrand
2020-02-03 17:03 ` Michael S. Tsirkin
2020-02-03 20:32 ` Tyler Sanderson
2020-02-03 21:22 ` Alexander Duyck
2020-02-03 23:16 ` Tyler Sanderson
2020-02-04 0:10 ` Alexander Duyck
2020-02-04 5:45 ` Michael S. Tsirkin
2020-02-04 8:29 ` David Hildenbrand
2020-02-04 18:52 ` Tyler Sanderson
2020-02-04 18:56 ` Michael S. Tsirkin
2020-02-04 19:17 ` David Hildenbrand
2020-02-04 23:58 ` Tyler Sanderson
2020-02-05 0:15 ` Tyler Sanderson
2020-02-05 6:57 ` Michael S. Tsirkin
2020-02-05 19:01 ` Tyler Sanderson
2020-02-05 19:22 ` Alexander Duyck
2020-02-05 21:44 ` Tyler Sanderson
2020-02-06 11:00 ` David Hildenbrand
2020-02-03 22:50 ` Nadav Amit
2020-02-04 8:35 ` David Hildenbrand
2020-02-04 8:40 ` Michael S. Tsirkin
2020-02-04 8:48 ` David Hildenbrand
2020-02-04 14:30 ` David Hildenbrand
2020-02-04 16:50 ` Michael S. Tsirkin
2020-02-04 16:56 ` David Hildenbrand
2020-02-04 20:33 ` Michael S. Tsirkin
2020-02-05 8:31 ` David Hildenbrand
2020-02-05 6:52 ` Wang, Wei W
2020-02-05 7:05 ` Michael S. Tsirkin
2020-02-05 8:50 ` Wang, Wei W
2020-02-05 6:49 ` Wang, Wei W
2020-02-05 8:19 ` David Hildenbrand
2020-02-05 8:54 ` Wang, Wei W
2020-02-05 8:56 ` David Hildenbrand
2020-02-05 9:00 ` Wang, Wei W
2020-02-05 9:05 ` David Hildenbrand
2020-02-05 9:19 ` Wang, Wei W
2020-02-05 9:22 ` David Hildenbrand
2020-02-05 9:35 ` Wang, Wei W
2020-02-05 9:37 ` David Hildenbrand
2020-02-05 9:49 ` Wang, Wei W
2020-02-05 9:58 ` David Hildenbrand
2020-02-05 10:25 ` Michael S. Tsirkin
2020-02-05 10:42 ` David Hildenbrand
2020-02-05 9:35 ` Michael S. Tsirkin
2020-02-05 18:43 ` Tyler Sanderson
2020-02-06 9:30 ` Wang, Wei W
2020-02-05 7:35 ` Nadav Amit
2020-02-05 8:19 ` David Hildenbrand
2020-02-05 10:27 ` Michael S. Tsirkin
2020-02-05 10:43 ` David Hildenbrand
2020-01-30 22:46 ` Tyler Sanderson [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to='CAJuQAmq3xPb5k=egCHwwYD=so8hwOCf6d42mUvVXOh2iLJ64dg@mail.gmail.com' \
--to=tysand@google.com \
--cc=david@redhat.com \
--cc=linux-mm@kvack.org \
--cc=mhocko@kernel.org \
--cc=mst@redhat.com \
--cc=rientjes@google.com \
--cc=virtualization@lists.linux-foundation.org \
--cc=wei.w.wang@intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
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).