From: Andrew W Elble <aweits@rit.edu>
To: cluster-devel.redhat.com
Subject: [Cluster-devel] [GFS2 PATCH v2 14/15] GFS2: Hold onto iopen glock longer when dinode creation fails
Date: Fri, 16 Oct 2015 09:49:02 -0400 [thread overview]
Message-ID: <m2vba6c4a9.fsf@discipline.rit.edu> (raw)
In-Reply-To: <820524051.47443676.1445002403265.JavaMail.zimbra@redhat.com> (Bob Peterson's message of "Fri, 16 Oct 2015 09:33:23 -0400")
> Hi Andy,
>
> Thanks. I'll investigate it.
>
> BTW, I haven't found any more blatant bugs during testing, however I'm
> debugging another issue. It seems as if I still have a reference counter
> issue somewhere because if I slam GFS2 hard enough, I can get it to
> accumulate millions of glocks that are never freed (unless memory pressure
> causes the glock shrinker to be called). Having the extra glocks is causing
> undue strain on the dlm, to the point where dlm can't keep up.
> So it seems likely that I'll do another revision here anyway.
> I'll keep you posted.
Bob,
Would that resemble something like this?
S glocks nondisk inode rgrp iopen flock quota j;rnl Total
S --------- ------- -------- ------- ------- ------- ----- ---- --------
S Unlocked: 1 67677 304 0 0 460 0 68442
S Locked: 2 1638575 630 1637848 0 780 1 3277836
S Total: 3 1706252 934 1637848 0 1240 1 3346278
S
S Held EX: 0 1 0 0 0 0 1 2
S Held SH: 1 1 0 1637847 0 0 0 1637849
S Held DF: 0 0 0 0 0 0 0 0
S G Waiting: 0 0 0 0 0 0 0 0
S P Waiting: 0 0 0 0 0 0 0 0
S DLM wait: 0
Thanks,
Andy
--
Andrew W. Elble
aweits at discipline.rit.edu
Infrastructure Engineer, Communications Technical Lead
Rochester Institute of Technology
PGP: BFAD 8461 4CCF DC95 DA2C B0EB 965B 082E 863E C912
next prev parent reply other threads:[~2015-10-16 13:49 UTC|newest]
Thread overview: 20+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-10-06 19:02 [Cluster-devel] [GFS2 PATCH v2 00/15] Fifteen patches related to file unlink->delete->new Bob Peterson
2015-10-06 19:02 ` [Cluster-devel] [GFS2 PATCH v2 01/15] GFS2: Update master statfs buffer with sd_statfs_spin locked Bob Peterson
2015-10-06 19:02 ` [Cluster-devel] [GFS2 PATCH v2 02/15] GFS2: Allow fail_gunlock3 to set the free_vfs_inode bit Bob Peterson
2015-10-06 19:02 ` [Cluster-devel] [GFS2 PATCH v2 03/15] GFS2: Protect log tail calculations with inside locks Bob Peterson
2015-10-06 19:02 ` [Cluster-devel] [GFS2 PATCH v2 04/15] GFS2: Wait for iopen glock dequeues Bob Peterson
2015-10-06 19:02 ` [Cluster-devel] [GFS2 PATCH v2 05/15] GFS2: Reintroduce a timeout in function gfs2_gl_hash_clear Bob Peterson
2015-10-06 19:02 ` [Cluster-devel] [GFS2 PATCH v2 06/15] GFS2: Prevent gl_delete work for re-used inodes Bob Peterson
2015-10-06 19:02 ` [Cluster-devel] [GFS2 PATCH v2 07/15] GFS2: Truncate address space mapping when deleting an inode Bob Peterson
2015-10-06 19:02 ` [Cluster-devel] [GFS2 PATCH v2 08/15] GFS2: Don't filter out I_FREEING inodes anymore Bob Peterson
2015-10-06 19:02 ` [Cluster-devel] [GFS2 PATCH v2 09/15] GFS2: generalize gfs2_check_blk_type Bob Peterson
2015-10-06 19:02 ` [Cluster-devel] [GFS2 PATCH v2 10/15] GFS2: Change from tr_touched to tr_bufs Bob Peterson
2015-10-06 19:02 ` [Cluster-devel] [GFS2 PATCH v2 11/15] GFS2: Add new function gfs2_inode_lookup_for_del Bob Peterson
2015-10-06 19:02 ` [Cluster-devel] [GFS2 PATCH v2 12/15] gfs2: Remove unused param non_block from gfs2_inode_lookup Bob Peterson
2015-10-06 19:02 ` [Cluster-devel] [GFS2 PATCH v2 13/15] gfs2: Use new variable i_gl instead of ip->i_gl Bob Peterson
2015-10-06 19:02 ` [Cluster-devel] [GFS2 PATCH v2 14/15] GFS2: Hold onto iopen glock longer when dinode creation fails Bob Peterson
2015-10-16 13:26 ` Andrew W Elble
2015-10-16 13:33 ` Bob Peterson
2015-10-16 13:49 ` Andrew W Elble [this message]
2015-10-16 14:02 ` Bob Peterson
2015-10-06 19:02 ` [Cluster-devel] [GFS2 PATCH v2 15/15] GFS2: Rework gfs2_evict_inode to prevent collisions with openers Bob Peterson
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=m2vba6c4a9.fsf@discipline.rit.edu \
--to=aweits@rit.edu \
/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).