Over the last few weeks we have been in the process of moving all of our
RHEL5 systems from 5.1 to 5.2.  Most of these upgrades have gone very
well, however, we have 4 or 5 systems which have spontaneously
"re-enabled" SELinux as part of the upgrade.  It appears that the
upgrade process is silently replacing the /etc/selinux/config file,
which we had modified to set SELINUX=disabled, with the default file
which has SELINUX=enforcing.

On reboot the systems fail to boot with various SELinux errors and a
warning that the file system needs to be relabled and the system is
rebooting.  Without manual intervention it is stuck in this loop.  We
can recover by adding "enforcing=0" to the kernel boot line and either
changing the file back to SELINUX=disabled or, if we decide to give
SElinux another spin, relabeling the filesystem, but I'm curious if
anyone else has seen this issue.

We have a mix of systems and about half have SELinux enabled, and the
other half do not, but this issue has only affected a handful of the
systems where SELinux is currently disabled.  Several other systems with
SELinux disabled upgraded without any issues.  The upgrades were
performed with a simple "yum update" not a CD/DVD upgrade.

Obviously we can recover from this issue without a major problem, but
I'm curious if others have seen it because we simply can't explain why
it seems to happen "randomly" rather than on every system that has
SELinux disabled.  We think it may be on systems that were previously
upgraded from RHEL4 where we always disabled SELinux.  We're really just
trying to find a pattern and determine if it's worth opening a case with
Redhat.

Later,
Tom

_______________________________________________
rhelv5-list mailing list
[email protected]
https://www.redhat.com/mailman/listinfo/rhelv5-list

Reply via email to