All of lore.kernel.org
 help / color / mirror / Atom feed
From: bzzz.tomas at gmail.com <bzzz.tomas@gmail.com>
To: lustre-devel@lists.lustre.org
Subject: [Lustre-devel] Queries regarding LDLM_ENQUEUE
Date: Wed, 20 Oct 2010 20:49:06 +0400	[thread overview]
Message-ID: <4CBF1D82.60508@gmail.com> (raw)
In-Reply-To: <4CBF1C42.1090109@psc.edu>

On 10/20/10 8:43 PM, Paul Nowoczynski wrote:
> It's for scalability reasons. When N clients traverse the namespace with
> the purpose of opening the same file the result is a storm of RPC
> requests which bear down on the metadata server. This type of activity
> becomes prohibitive especially when you start considering client counts
>  > 10^4. An operation such as this is ripe for optimization because
> every client in the network is trying to build the same state. If you
> have a method for a single client to 'learn' the final state, i.e. the
> pathname -> fid translation, and broadcast it to its cohorts, it's a
> huge win because it eliminates an O(N) operation.
> paul

clear enough, but what is the bottleneck here: MDS to handle lots of
RPCs or network to pass RPCs ?

thanks, z

  reply	other threads:[~2010-10-20 16:49 UTC|newest]

Thread overview: 35+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2010-10-18 23:33 [Lustre-devel] Queries regarding LDLM_ENQUEUE Vilobh Meshram
2010-10-19 15:46 ` Fan Yong
2010-10-19 20:28 ` Vilobh Meshram
2010-10-19 22:53   ` Andreas Dilger
2010-10-20  2:04     ` Vilobh Meshram
2010-10-20  7:55       ` Andreas Dilger
2010-10-20  8:11         ` bzzz.tomas at gmail.com
2010-10-20  8:24           ` Andreas Dilger
2010-10-20  8:30             ` bzzz.tomas at gmail.com
2010-10-20  8:38               ` Nikita Danilov
2010-10-20 14:45                 ` Nicolas Williams
2010-10-20 13:30               ` Eric Barton
2010-10-20 13:40                 ` bzzz.tomas at gmail.com
2010-10-20 14:51                 ` Paul Nowoczynski
2010-10-20 14:55                   ` Nicolas Williams
2010-10-20 15:16                     ` Paul Nowoczynski
2010-10-20 16:07                       ` Andreas Dilger
2010-10-20 15:22                   ` bzzz.tomas at gmail.com
2010-10-20 16:43                     ` Paul Nowoczynski
2010-10-20 16:49                       ` bzzz.tomas at gmail.com [this message]
2010-10-20 17:11                         ` Paul Nowoczynski
2010-10-20 17:18                           ` bzzz.tomas at gmail.com
2010-10-20 17:25                             ` Paul Nowoczynski
2010-10-20 17:27                             ` Andreas Dilger
2010-10-20 17:29                             ` Nicolas Williams
2010-10-20 17:40                               ` bzzz.tomas at gmail.com
2010-10-20 18:01                                 ` Andreas Dilger
2010-10-20 18:09                                   ` bzzz.tomas at gmail.com
2010-10-20 16:35                 ` Andreas Dilger
2010-10-20 16:46                   ` Paul Nowoczynski
2010-10-20 17:00                     ` Andreas Dilger
2010-10-20 17:13                       ` Nicolas Williams
2010-10-20 17:30                         ` Andreas Dilger
2010-10-20 17:01                     ` Nicolas Williams
2010-10-22  2:33         ` Vilobh Meshram

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=4CBF1D82.60508@gmail.com \
    --to=bzzz.tomas@gmail.com \
    --cc=lustre-devel@lists.lustre.org \
    /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.