[bug-notifications] [findutils 0004076]: gupdatedb should check if it's still running
Mantis Bug Tracker
noreply at opencsw.org
Sat May 14 14:59:55 CEST 2011
A NOTE has been added to this issue.
======================================================================
https://www.opencsw.org/mantis/view.php?id=4076
======================================================================
Reported By: holzi
Assigned To:
======================================================================
Project: findutils
Issue ID: 4076
Category: regular use
Reproducibility: sometimes
Severity: tweak
Priority: normal
Status: new
======================================================================
Date Submitted: 2009-12-10 10:58 CET
Last Modified: 2011-05-14 14:59 CEST
======================================================================
Summary: gupdatedb should check if it's still running
Description:
I didn't check the whole script but I looks like there is no check.
Reason: When the gfind stuff takes longer then a day. ( e.g. somtimes on
the buildfram because it's called by every zone) It's started again. Which
could have bad effects on the system performance.
======================================================================
----------------------------------------------------------------------
(0009047) jay (reporter) - 2011-05-14 14:59
https://www.opencsw.org/mantis/view.php?id=4076#c9047
----------------------------------------------------------------------
You can just mutually exclude multiple runs by using "lockfile -l 60" or
similar.
If the problem is NFS performance, then you can build a locate database on
the NFS server and have the clients all search that by using LOCATEPATH (as
long as all the clients mount the exported filesystem in the same place).
Then you just add nfs to the list of filesystems updatedb doesn't search.
More information about the bug-notifications
mailing list