From: Keir Fraser <keir.fraser@eu.citrix.com>
To: Brendan Cully <brendan@cs.ubc.ca>,
Ian Jackson <Ian.Jackson@eu.citrix.com>
Cc: "xen-devel@lists.xensource.com" <xen-devel@lists.xensource.com>,
Andreas Olsowski <andreas.olsowski@uni.leuphana.de>,
"Zhai, Edwin" <edwin.zhai@intel.com>
Subject: Re: slow live magration / xc_restore on xen4 pvops
Date: Thu, 3 Jun 2010 16:18:39 +0100 [thread overview]
Message-ID: <C82D865F.169ED%keir.fraser@eu.citrix.com> (raw)
In-Reply-To: <20100603150305.GA53591@zanzibar.domain.invalid>
On 03/06/2010 16:03, "Brendan Cully" <brendan@cs.ubc.ca> wrote:
> I see no evidence that Remus has anything to do with the live
> migration performance regression discussed in this thread, and I
> haven't seen any other reported issues either. I think the mlock issue
> is a much more likely candidate.
I agree it's probably lack of batching plus expensive mlocks. The
performance difference between different machines under test is either
because one runs out of 2MB superpage extents before the other (for some
reason) or because mlock operations are for some reason much more likely to
take a slow path in the kernel (possibly including disk i/o) for some
reason.
We need to get batching back, and Edwin is on the case for that: I hope
Andreas will try out Edwin's patch to work towards that. We can also reduce
mlock cost by mlocking some domain_restore arrays across the entire restore
operation, I should imagine.
-- Keir
next prev parent reply other threads:[~2010-06-03 15:18 UTC|newest]
Thread overview: 32+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-06-01 17:49 XCP AkshayKumar Mehta
2010-06-01 19:06 ` XCP Jonathan Ludlam
2010-06-01 19:15 ` XCP AkshayKumar Mehta
2010-06-03 3:03 ` XCP AkshayKumar Mehta
2010-06-03 10:24 ` XCP Jonathan Ludlam
2010-06-03 17:20 ` XCP AkshayKumar Mehta
2010-08-31 1:33 ` XCP - iisues with XCP .5 AkshayKumar Mehta
2010-06-01 21:17 ` slow live magration / xc_restore on xen4 pvops Andreas Olsowski
2010-06-02 7:11 ` Keir Fraser
2010-06-02 15:46 ` Andreas Olsowski
2010-06-02 15:55 ` Keir Fraser
2010-06-02 16:18 ` Ian Jackson
2010-06-02 16:20 ` Ian Jackson
2010-06-02 16:24 ` Keir Fraser
2010-06-03 1:04 ` Brendan Cully
2010-06-03 4:31 ` Brendan Cully
2010-06-03 5:47 ` Keir Fraser
2010-06-03 6:45 ` Brendan Cully
2010-06-03 6:53 ` Jeremy Fitzhardinge
2010-06-03 6:55 ` Brendan Cully
2010-06-03 7:12 ` Keir Fraser
2010-06-03 8:58 ` Zhai, Edwin
2010-06-09 13:32 ` Keir Fraser
2010-06-02 16:27 ` Brendan Cully
2010-06-03 10:01 ` Ian Jackson
2010-06-03 15:03 ` Brendan Cully
2010-06-03 15:18 ` Keir Fraser [this message]
2010-06-03 17:15 ` Ian Jackson
2010-06-03 17:29 ` Brendan Cully
2010-06-03 18:02 ` Ian Jackson
2010-06-02 22:59 ` Andreas Olsowski
2010-06-10 9:27 ` Keir Fraser
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=C82D865F.169ED%keir.fraser@eu.citrix.com \
--to=keir.fraser@eu.citrix.com \
--cc=Ian.Jackson@eu.citrix.com \
--cc=andreas.olsowski@uni.leuphana.de \
--cc=brendan@cs.ubc.ca \
--cc=edwin.zhai@intel.com \
--cc=xen-devel@lists.xensource.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).