Date: Mon, 22 Aug 2016 12:10:16 -0400 From: Robert Fitzpatrick <robert@webtent.org> To: David Christensen <dpchrist@holgerdanske.com> Cc: freebsd-questions@freebsd.org Subject: Re: Monitoring server for crashes Message-ID: <57BB23E8.10906@webtent.org> In-Reply-To: <f27d9894-ebae-5b52-4a98-8d9c812559bf@holgerdanske.com> References: <57ADDA5F.4000405@webtent.org> <61294.128.135.52.6.1471013465.squirrel@cosmo.uchicago.edu> <57ADF096.8010608@webtent.org> <11590.128.135.52.6.1471018231.squirrel@cosmo.uchicago.edu> <f27d9894-ebae-5b52-4a98-8d9c812559bf@holgerdanske.com>
next in thread | previous in thread | raw e-mail | index | archive | help
David Christensen wrote: > I would advise testing before changing anything. The strategy is: > devise a reproducible test that invokes the bug, use the test to isolate > the bug, fix the bug, re-run test to verify the bug is fixed, re-run the > test periodically to verify that that bug has not returned. Sorry so long to post back, yes, I decided this approach was best and found full backup of the PostgreSQL database causes crash while backing up individual databases (including the largest) does not. Perhaps Postgres just provides the environment to trigger the crash and might not be the culprit, but it's the only way I can reproduce at the moment. I did find the IBM specs for memory of that board and have not compared yet to the box, maybe soon. Just slammed with work and this box is replicated. While I don't want it down, I can take my time finding the real solution if workaround exists. If memory is not right, will change and re-run the full backup of postgres. This box is currently running FreeBSD 10.0-RELEASE-p18. -- Robert
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?57BB23E8.10906>