Description of problem: Postgres 8.1.11 as shipped with RHEL5 does not work with the postgres-8.sh / postgres-8.metadata files. (1) [FEATURE] The default config file is given, but the default data directory should also be given. (2) [BUG] Postmaster 8.1.11 looks for ip_addresses, not ip_address in postgresql.conf. This causes postmaster to fail to start. (3) [BUG] The PID file directory, /var/run/cluster/postgres-8, is owned by root/root. This makes it impossible for the postmaster program (running as 'postgres' typically) to create a PID file there. A solution is to create the file but leave it empty prior to starting postmaster & changing the ownership of the file to the postgres user. (4) [BUG] post master doesn't fork; we should spawn it in the background
one might also unlink the PID file if it was empty in stop_generic()
Created attachment 317213 [details] Patch
This request was evaluated by Red Hat Product Management for inclusion in a Red Hat Enterprise Linux maintenance release. Product Management has requested further review of this request by Red Hat Engineering, for potential inclusion in a Red Hat Enterprise Linux Update release for currently deployed products. This request is not yet committed for inclusion in an Update release.
This bugzilla has Keywords: Regression. Since no regressions are allowed between releases, it is also being proposed as a blocker for this release. Please resolve ASAP.
Created attachment 325029 [details] patch to make postgres-8.sh use su instead of sudo I have an additional problem with this package wrt postgres-8.sh, that is not corrected by the previous patch. The "sudo" statement doesn't work, it complains (once I have redirected to a file instead of /dev/null) about needing a pty. Replacing sudo by su -c works (see patch). Why use sudo in the first place? It's dependent upon a working /etc/sudoers while su should work all the time for root. sudo-1.6.8p12-12.el5.x86_64.rpm rgmanager-2.0.43-1.el5.x86_64.rpm Note that I'm using Postgres 8.3.3 (upstream), but this bug doesn't look like something caused by postgres, does it?
Addendum: the script is rendered quite fragile by the fact that it fails to start if a pid file exists (even zero-length). This needs to be fixed.
http://git.fedorahosted.org/git/?p=cluster.git;a=commit;h=116b8d72fdde451a53838b21e4b713e05b822016
Without changing sudo to su as described, rgmanager "successfully" starts the postgres application when actually it's not running at all.
(In reply to comment #8) > Addendum: the script is rendered quite fragile by the fact that it fails to > start if a pid file exists (even zero-length). This needs to be fixed. Agreed. However, this is a different issue which affects config-utils.sh generally.
(In reply to comment #12) > (In reply to comment #8) > > Addendum: the script is rendered quite fragile by the fact that it fails to > > start if a pid file exists (even zero-length). This needs to be fixed. > > Agreed. However, this is a different issue which affects config-utils.sh > generally. https://bugzilla.redhat.com/show_bug.cgi?id=474444 <-- pid file issue
An advisory has been issued which should help the problem described in this bug report. This report is therefore being closed with a resolution of ERRATA. For more information on therefore solution and/or where to find the updated files, please follow the link below. You may reopen this bug report if the solution does not work for you. http://rhn.redhat.com/errata/RHBA-2009-0101.html