8.0.0 on disk fails regress test, but on ram disk it is fine...

From: Jeff Ross <jross(at)openvistas(dot)net>
To: pgsql-bugs(at)postgresql(dot)org
Subject: 8.0.0 on disk fails regress test, but on ram disk it is fine...
Date: 2005-01-22 23:45:33
Message-ID: B4C1C27C-6CCF-11D9-85D5-000393B4C506@openvistas.net
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-bugs pgsql-docs pgsql-www

I'm preparing to upgrade our PostgreSQL installation from 7.3.5 to
8.0.0. I downloaded, configured, compiled, and did gmake check and
much to my surprise, 26 of the 96 tests failed.

Examination of regression.diffs shows that 1 of the failures is random
(and it is ignored) but 23 of the 26 errors are:

! psql: could not send startup packet: Broken pipe

Interestingly enough, if I load the source code onto a 1GB ram disk and
configure, compile and check it there, all 96 tests pass.

Hardware is RAID 1 using 2 disks with one additional disk as a hot
spare, dual 2.66 GHz Xeon processors with 2 GB of RAM. Operating
system is OpenBSD 3.6.

I'm a little leary of upgrading my PostgreSQL installation with those
kind of failure rates.

I'd appreciate any insight as to why so many tests would fail when run
on disk and pass when run in ram.

Thanks,

Jeff Ross

In response to

Browse pgsql-bugs by date

  From Date Subject
Next Message Tom Lane 2005-01-23 00:09:21 Re: BUG #1426: pg_restore SIGSEGV on 7.2.2 dump
Previous Message Bruce Momjian 2005-01-22 23:01:11 Re: BUG #1414: DOC - pl/Perl hash tags missing

Browse pgsql-docs by date

  From Date Subject
Next Message Tom Lane 2005-01-23 00:40:06 Re: BUG #1414: DOC - pl/Perl hash tags missing
Previous Message Bruce Momjian 2005-01-22 23:01:11 Re: BUG #1414: DOC - pl/Perl hash tags missing

Browse pgsql-www by date

  From Date Subject
Next Message Tom Lane 2005-01-23 00:40:06 Re: BUG #1414: DOC - pl/Perl hash tags missing
Previous Message Oleg Bartunov 2005-01-22 23:35:19 Re: request: search engine for PostgreSQL site