From: Chris Webb <chris@arachsys.com>
To: Avi Kivity <avi@redhat.com>
Cc: balbir@linux.vnet.ibm.com,
KVM development list <kvm@vger.kernel.org>,
Rik van Riel <riel@surriel.com>,
KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>,
"linux-mm@kvack.org" <linux-mm@kvack.org>,
"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH][RF C/T/D] Unmapped page cache control - via boot parameter
Date: Mon, 15 Mar 2010 20:23:54 +0000 [thread overview]
Message-ID: <20100315202353.GJ3840@arachsys.com> (raw)
In-Reply-To: <4B9DEF81.6020802@redhat.com>
Avi Kivity <avi@redhat.com> writes:
> On 03/15/2010 10:07 AM, Balbir Singh wrote:
>
> >Yes, it is a virtio call away, but is the cost of paying twice in
> >terms of memory acceptable?
>
> Usually, it isn't, which is why I recommend cache=off.
Hi Avi. One observation about your recommendation for cache=none:
We run hosts of VMs accessing drives backed by logical volumes carved out
from md RAID1. Each host has 32GB RAM and eight cores, divided between (say)
twenty virtual machines, which pretty much fill the available memory on the
host. Our qemu-kvm is new enough that IDE and SCSI drives with writeback
caching turned on get advertised to the guest as having a write-cache, and
FLUSH gets translated to fsync() by qemu. (Consequently cache=writeback
isn't acting as cache=neverflush like it would have done a year ago. I know
that comparing performance for cache=none against that unsafe behaviour
would be somewhat unfair!)
Wasteful duplication of page cache between guest and host notwithstanding,
turning on cache=writeback is a spectacular performance win for our guests.
For example, even IDE with cache=writeback easily beats virtio with
cache=none in most of the guest filesystem performance tests I've tried. The
anecdotal feedback from clients is also very strongly in favour of
cache=writeback.
With a host full of cache=none guests, IO contention between guests is
hugely problematic with non-stop seek from the disks to service tiny
O_DIRECT writes (especially without virtio), many of which needn't have been
synchronous if only there had been some way for the guest OS to tell qemu
that. Running with cache=writeback seems to reduce the frequency of disk
flush per guest to a much more manageable level, and to allow the host's
elevator to optimise writing out across the guests in between these flushes.
Cheers,
Chris.
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
next prev parent reply other threads:[~2010-03-15 20:28 UTC|newest]
Thread overview: 49+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-03-15 7:22 [PATCH][RF C/T/D] Unmapped page cache control - via boot parameter Balbir Singh
2010-03-15 7:48 ` Avi Kivity
2010-03-15 8:07 ` Balbir Singh
2010-03-15 8:27 ` Avi Kivity
2010-03-15 9:17 ` Balbir Singh
2010-03-15 9:27 ` Avi Kivity
2010-03-15 10:45 ` Balbir Singh
2010-03-15 18:48 ` Anthony Liguori
2010-03-16 9:05 ` Avi Kivity
2010-03-19 7:23 ` Dave Hansen
2010-03-15 20:23 ` Chris Webb [this message]
2010-03-15 23:43 ` Anthony Liguori
2010-03-16 0:43 ` Christoph Hellwig
2010-03-16 1:27 ` Anthony Liguori
2010-03-16 8:19 ` Christoph Hellwig
2010-03-17 15:14 ` Chris Webb
2010-03-17 15:55 ` Anthony Liguori
2010-03-17 16:27 ` Chris Webb
2010-03-22 21:04 ` Chris Webb
2010-03-22 21:07 ` Avi Kivity
2010-03-22 21:10 ` Chris Webb
2010-03-17 16:27 ` Balbir Singh
2010-03-17 17:05 ` Vivek Goyal
2010-03-17 19:11 ` Chris Webb
2010-03-16 3:16 ` Balbir Singh
2010-03-16 9:17 ` Avi Kivity
2010-03-16 9:54 ` Kevin Wolf
2010-03-16 10:16 ` Avi Kivity
2010-03-16 10:26 ` Christoph Hellwig
2010-03-16 10:36 ` Avi Kivity
2010-03-16 10:44 ` Christoph Hellwig
2010-03-16 11:08 ` Avi Kivity
2010-03-16 14:27 ` Balbir Singh
2010-03-16 15:59 ` Avi Kivity
2010-03-17 8:49 ` Christoph Hellwig
2010-03-17 9:10 ` Avi Kivity
2010-03-17 15:24 ` Chris Webb
2010-03-17 16:22 ` Avi Kivity
2010-03-17 16:40 ` Avi Kivity
2010-03-17 16:47 ` Chris Webb
2010-03-17 16:53 ` Avi Kivity
2010-03-17 16:58 ` Christoph Hellwig
2010-03-17 17:03 ` Avi Kivity
2010-03-17 16:57 ` Christoph Hellwig
2010-03-17 17:06 ` Avi Kivity
2010-03-17 16:52 ` Christoph Hellwig
2010-03-17 17:02 ` Avi Kivity
2010-03-15 15:46 ` Randy Dunlap
2010-03-16 3:21 ` Balbir Singh
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=20100315202353.GJ3840@arachsys.com \
--to=chris@arachsys.com \
--cc=avi@redhat.com \
--cc=balbir@linux.vnet.ibm.com \
--cc=kamezawa.hiroyu@jp.fujitsu.com \
--cc=kvm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=riel@surriel.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).