From: Peter Zijlstra <peterz@infradead.org>
To: Nikhil Rao <ncrao@google.com>
Cc: Ingo Molnar <mingo@elte.hu>, Mike Galbraith <efault@gmx.de>,
linux-kernel@vger.kernel.org,
"Nikunj A. Dadhania" <nikunj@linux.vnet.ibm.com>,
Srivatsa Vaddagiri <vatsa@linux.vnet.ibm.com>,
Stephan Barwolf <stephan.baerwolf@tu-ilmenau.de>
Subject: Re: [PATCH v1 00/19] Increase resolution of load weights
Date: Fri, 13 May 2011 09:19:47 +0200 [thread overview]
Message-ID: <1305271187.17430.8.camel@twins> (raw)
In-Reply-To: <BANLkTimsvvWqcGanJs688hytLx1_jb0itw@mail.gmail.com>
On Thu, 2011-05-12 at 10:30 -0700, Nikhil Rao wrote:
> On Thu, May 12, 2011 at 2:08 AM, Peter Zijlstra <peterz@infradead.org> wrote:
> > On Thu, 2011-05-05 at 18:29 -0700, Nikhil Rao wrote:
> >> > It's a cost/benefit analysis and for 32-bit systems the benefits seem to be
> >> > rather small, right?
> >> >
> >>
> >> Yes, that's right. The benefits for 32-bit systems do seem to be limited.
> >
> > deep(er) hierarchies on 32 bits still require this, it would be good to
> > verify that the cgroup mess created by the insanity called libvirt will
> > indeed work as expected.
> >
>
> I went through the libvirt docs and from what I understand, it creates
> a hierarchy which is about 3 levels deep and has as many leaf nodes as
> guest VMs.
That sounds about right with what I remember people telling me
earlier ;-)
> Taking this graphic from
> http://berrange.com/posts/2009/12/03/using-cgroups-with-libvirt-and-lxckvm-guests-in-fedora-12/
>
> $ROOT
> |
> +- libvirt (all virtual machines/containers run by libvirtd)
> |
> +- lxc (all LXC containers run by libvirtd)
> | |
> | +- guest1 (LXC container called 'guest1')
> | +- guest2 (LXC container called 'guest2')
> | +- guest3 (LXC container called 'guest3')
> | +- ... (LXC container called ...)
> |
> +- qemu (all QEMU/KVM containers run by libvirtd)
> |
> +- guest1 (QENU machine called 'guest1')
> +- guest2 (QEMU machine called 'guest2')
> +- guest3 (QEMU machine called 'guest3')
> +- ... (QEMU machine called ...)
>
> Assuming the tg shares given to libvirt, lxc and qemu containers are
> the defaults, the load balancer should be able to deal with the
> current resolution on 32-bit. Back of the envelope calculations using
> that approach I mentioned earlier (i.e. log_b(1024/NR_CPU)) says you
> need > 64 VMs before you run out of resolution. I think that might be
> too much to expect from a 8-cpu 32-bit machine ;-)
Quite so, get a real machine etc.. ;-) But then, there's always some
weird people out there, but I think we can tell them to run a 64bit
kernel.
prev parent reply other threads:[~2011-05-13 7:20 UTC|newest]
Thread overview: 34+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-05-02 1:18 [PATCH v1 00/19] Increase resolution of load weights Nikhil Rao
2011-05-02 1:18 ` [PATCH v1 01/19] sched: introduce SCHED_POWER_SCALE to scale cpu_power calculations Nikhil Rao
2011-05-02 1:19 ` [PATCH v1 02/19] sched: increase SCHED_LOAD_SCALE resolution Nikhil Rao
2011-05-02 1:19 ` [PATCH v1 03/19] sched: use u64 for load_weight fields Nikhil Rao
2011-05-02 1:19 ` [PATCH v1 04/19] sched: update cpu_load to be u64 Nikhil Rao
2011-05-02 1:19 ` [PATCH v1 05/19] sched: update this_cpu_load() to return u64 value Nikhil Rao
2011-05-02 1:19 ` [PATCH v1 06/19] sched: update source_load(), target_load() and weighted_cpuload() to use u64 Nikhil Rao
2011-05-02 1:19 ` [PATCH v1 07/19] sched: update find_idlest_cpu() to use u64 for load Nikhil Rao
2011-05-02 1:19 ` [PATCH v1 08/19] sched: update find_idlest_group() to use u64 Nikhil Rao
2011-05-02 1:19 ` [PATCH v1 09/19] sched: update division in cpu_avg_load_per_task to use div_u64 Nikhil Rao
2011-05-02 1:19 ` [PATCH v1 10/19] sched: update wake_affine path to use u64, s64 for weights Nikhil Rao
2011-05-02 1:19 ` [PATCH v1 11/19] sched: update update_sg_lb_stats() to use u64 Nikhil Rao
2011-05-02 1:19 ` [PATCH v1 12/19] sched: Update update_sd_lb_stats() " Nikhil Rao
2011-05-02 1:19 ` [PATCH v1 13/19] sched: update f_b_g() to use u64 for weights Nikhil Rao
2011-05-02 1:19 ` [PATCH v1 14/19] sched: change type of imbalance to be u64 Nikhil Rao
2011-05-02 1:19 ` [PATCH v1 15/19] sched: update h_load to use u64 Nikhil Rao
2011-05-02 1:19 ` [PATCH v1 16/19] sched: update move_task() and helper functions to use u64 for weights Nikhil Rao
2011-05-02 1:19 ` [PATCH v1 17/19] sched: update f_b_q() to use u64 for weighted cpuload Nikhil Rao
2011-05-02 1:19 ` [PATCH v1 18/19] sched: update shares distribution to use u64 Nikhil Rao
2011-05-02 1:19 ` [PATCH v1 19/19] sched: convert atomic ops in shares update to use atomic64_t ops Nikhil Rao
2011-05-02 6:14 ` [PATCH v1 00/19] Increase resolution of load weights Ingo Molnar
2011-05-04 0:58 ` Nikhil Rao
2011-05-04 1:07 ` Nikhil Rao
2011-05-04 11:13 ` Ingo Molnar
2011-05-06 1:29 ` Nikhil Rao
2011-05-06 6:59 ` Ingo Molnar
2011-05-11 0:14 ` Nikhil Rao
2011-05-11 6:59 ` Ingo Molnar
2011-05-12 8:56 ` Nikhil Rao
2011-05-12 10:55 ` Ingo Molnar
2011-05-12 18:44 ` Nikhil Rao
2011-05-12 9:08 ` Peter Zijlstra
2011-05-12 17:30 ` Nikhil Rao
2011-05-13 7:19 ` Peter Zijlstra [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=1305271187.17430.8.camel@twins \
--to=peterz@infradead.org \
--cc=efault@gmx.de \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@elte.hu \
--cc=ncrao@google.com \
--cc=nikunj@linux.vnet.ibm.com \
--cc=stephan.baerwolf@tu-ilmenau.de \
--cc=vatsa@linux.vnet.ibm.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.