7 Jun
2016
7 Jun
'16
8:37 a.m.
On 6 June 2016 at 11:46, Martin Perina <mperina@redhat.com> wrote:
No it's not, this is some strange failure in 3.5 upgrade script which seems to me impossible. The patch [1] didn't contain any db related stuff, so upgrade from master before this patch to master with this patch shouldn't fail (and I just verified locally that this upgrade works fine). So I assume this was some kind of mess in db content prior the upgrade job.
This is close to impossible, the upgrade job removes the postgresql package and deletes "/var/lib/pgsql/data" as can be seen here: https://gerrit.ovirt.org/gitweb?p=jenkins.git;a=blob;f=jobs/ovirt-engine_upg...