[colug-432] RHEL6 + Postgres + SELinux = Sadness

R P Herrold herrold at owlriver.com
Sun Apr 25 12:24:05 EDT 2010


On Sun, 25 Apr 2010, Joshua Kramer wrote:

> I have a bit of a quandary here.  I am using the RHEL6 64 bit beta and the
> PG 8.4.2 that comes with the distro.  Under normal circumstances,

I don't get it -- this is a beta -- unusual to their prior 
practice [perhaps in part from recognizing the 'facts of life' 
that that vendor may not be able to sell SLA backed service 
and updates, but at least they can prevent defections to 
Debian/Canonical LTS, or to Novell/SuSE, or even other *nix 
variants, if they can can have a 'relief valve' of CentOS to 
prevent such defections], it is a ** public ** beta, designed 
to permit the wider community of users of RHEL to use the 
designated mailing list, and the designated bugzilla 
component, to gather 'corner case' use cases and to permit the 
upstream to address the matters which end users find, that 
their ues matrix and testing do not expose.

Heaven know some in the CentOS camp wallow in reports in the C 
ML's of similar reports without mention of effectively 
reporting their issues already.  Some on the C QA, in the C 
DEV, and as it turns out, the wider C community seem to enjoy 
gossiping and kvetching about their upstream's usage dark 
corners.  Me? I'd rather report bugs in the proper venue, talk 
about previously unreported issues and posible resolutions in 
the venues where the reports will be seen, tallied, and 
fixes hopefully folded in and addressed by Gold release time

File a bug in Red Hat's bugzilla; participate on the Red Hat 
provided mailing list for the purpose of discussing the 
RHEL 6 beta bugs.

You already know that the 'permissive' use permits a 
workaround, and I am pretty sure I have seen mention on their 
'beta' ML for this candidate product of using the audit log, 
and the rule extraction tool, to create the additional local 
ruleset additions needed.

The upstream has consciously avoided using '/opt/ forever, but 
has provided the tools so that local variations may be 
accomodated

-- Russ herrold


More information about the colug-432 mailing list