From: George Dunlap <george.dunlap@eu.citrix.com>
To: Dulloor <dulloor@gmail.com>,
"xen-devel@lists.xensource.com" <xen-devel@lists.xensource.com>
Subject: Re: [PATCH 0 of 5] Add credit2 scheduler (EXPERIMENTAL)
Date: Thu, 15 Apr 2010 15:17:40 +0100 [thread overview]
Message-ID: <4BC72004.2030203@eu.citrix.com> (raw)
In-Reply-To: <h2x940bcfd21004140841kcdffe330xff5d749d43392fe3@mail.gmail.com>
Dulloor wrote:
> As we talked before, I am interested in improving the mutiple-socket
> scenario and adding the load balancing functionalilty, which could
> provide an acceptable alternative to pinning vcpus to sockets (for my
> NUMA work). I am going over your patch right now, but what are your
> thoughts ?
>
That would be great -- my focus for the next several months will be
setting up a testing infrastructure to automatically test performance of
different workloads mixes so I can hone the algorithm and test regressions.
My idea with load balancing was to do this:
* One runqueue per L2 cache.
* Add code to calculate the load of a runqueue. Load would be the
average (~integral) of (vcpus running + vcpus on runqueue). I was
planning on doing accurate load calculation, rather than sample-based,
and falling back to sample-based if accurate turned out to be too slow.
* Calculate the load contributed by various vcpus.
* At regular intervals, determine of some kind of balancing needs to be
done by looking at the overall runqueue load and placing based on
"contributory" load of each VCPU.
Does that make sense? Thoughts?
I have some old patches that calculated accurate load, I could dig them
up if you wanted something to start with. (I don't think they'll apply
cleanly at the moment.)
Thanks,
-George
> -dulloor
>
> On Wed, Apr 14, 2010 at 6:26 AM, George Dunlap
> <george.dunlap@eu.citrix.com> wrote:
>
>> This patch series introduces the credit2 scheduler. The first two patches
>> introduce changes necessary to allow the credit2 shared runqueue functionality
>> to work properly; the last two implement the functionality itself.
>>
>> The scheduler is still in the experimental phase. There's lots of
>> opportunity to contribute with independent lines of development; email
>> George Dunlap <george.dunlap@eu.citrix.com> or check out the wiki page
>> http://wiki.xensource.com/xenwiki/Credit2_Scheduler_Development for ideas
>> and status updates.
>>
>> 19 files changed, 1453 insertions(+), 21 deletions(-)
>> tools/libxc/Makefile | 1
>> tools/libxc/xc_csched2.c | 50 +
>> tools/libxc/xenctrl.h | 8
>> tools/python/xen/lowlevel/xc/xc.c | 58 +
>> tools/python/xen/xend/XendAPI.py | 3
>> tools/python/xen/xend/XendDomain.py | 54 +
>> tools/python/xen/xend/XendDomainInfo.py | 4
>> tools/python/xen/xend/XendNode.py | 4
>> tools/python/xen/xend/XendVMMetrics.py | 1
>> tools/python/xen/xend/server/SrvDomain.py | 14
>> tools/python/xen/xm/main.py | 82 ++
>> xen/arch/ia64/vmx/vmmu.c | 6
>> xen/common/Makefile | 1
>> xen/common/sched_credit.c | 8
>> xen/common/sched_credit2.c | 1125 +++++++++++++++++++++++++++++
>> xen/common/schedule.c | 22
>> xen/include/public/domctl.h | 4
>> xen/include/public/trace.h | 1
>> xen/include/xen/sched-if.h | 28
>>
>> _______________________________________________
>> Xen-devel mailing list
>> Xen-devel@lists.xensource.com
>> http://lists.xensource.com/xen-devel
>>
>>
next prev parent reply other threads:[~2010-04-15 14:17 UTC|newest]
Thread overview: 21+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-04-14 10:26 [PATCH 0 of 5] Add credit2 scheduler (EXPERIMENTAL) George Dunlap
2010-04-14 10:26 ` [PATCH 1 of 5] credit2: Add context_saved scheduler callback George Dunlap
2010-04-14 10:26 ` [PATCH 2 of 5] credit2: Flexible cpu-to-schedule-spinlock mappings George Dunlap
2010-04-14 10:26 ` [PATCH 3 of 5] credit2: Add a scheduler-specific schedule trace class George Dunlap
2010-04-14 10:26 ` [PATCH 4 of 5] credit2: Add credit2 scheduler to hypervisor George Dunlap
2010-04-14 10:26 ` [PATCH 5 of 5] credit2: Add toolstack options to control credit2 scheduler parameters George Dunlap
[not found] ` <7db7f696-1f0b-44d0-8f7b-eea1be5167dd@default>
2010-04-14 14:29 ` [PATCH 0 of 5] Add credit2 scheduler (EXPERIMENTAL) George Dunlap
2010-04-14 14:52 ` Keir Fraser
2010-04-14 15:59 ` Dan Magenheimer
2010-04-14 16:23 ` Keir Fraser
2010-04-14 16:31 ` Dulloor
2010-04-14 16:36 ` Keir Fraser
2010-04-14 17:04 ` Dan Magenheimer
2010-04-14 16:46 ` Dan Magenheimer
2010-04-15 20:11 ` Dan Magenheimer
[not found] ` <4BC664E1.7090304@purdue.edu>
2010-04-15 13:53 ` George Dunlap
2010-04-15 16:46 ` Naresh Rapolu
2010-04-15 17:33 ` Dulloor
2010-04-15 18:57 ` Naresh Rapolu
[not found] ` <h2x940bcfd21004140841kcdffe330xff5d749d43392fe3@mail.gmail.com>
2010-04-15 14:17 ` George Dunlap [this message]
2010-04-17 20:29 ` Dulloor
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=4BC72004.2030203@eu.citrix.com \
--to=george.dunlap@eu.citrix.com \
--cc=dulloor@gmail.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).