Created attachment 837692 [details] Error from console Description of problem: Provisioning of RHSS2.1 through use of kickstart file with latest GlusterFS rpms i.e. glusterfs-libs-3.4.0.50rhs-1.el6rhs.x86_64.rpm fails with below error in console. "Error Installing package A fatal error occurred when installing the glusterfs-libs package. This could indicate errors when reading the installation media. Installation can't continue" The issue can be seen when ISO "RHSS-2.1-20131215.n.0-RHS-x86_64-DVD1" or RHS2.1U1 GA ISO used as source tree This came in to light when provisioning of the RHSS machines failed for BVT runs. Beaker provisioning failed for the latest RPMs with below error. Installing glusterfs-libs-3.4.0.50rhs-1387276796.el6.x86_64 error: %pretrans(glusterfs-libs-3.4.0.50rhs-1387276796.el6.x86_64) scriptlet failed, exit status 127 This issue is not seen for "yum install" or "yum update" for mentioned gluster RPMs. But I think we will hit this issue when we will create an ISO with the latest RPMs. Version-Release number of selected component (if applicable): glusterfs-libs-3.4.0.50rhs-1.el6rhs.x86_64.rpm How reproducible: Always Steps to Reproduce: 1. Create a kickstart file for RHSS 2. In the kickstart file use the repo which has the latest RPMs 2. Use RHSS-2.1-20131215.n.0-RHS-x86_64-DVD1 as source tree and try to install the OS Actual results: Expected results: Additional info:
This is interesting issue - can we gather more information here? the 'pretrans' part doesn't seem to do much unless the 'pidof' part thinks that 'glusterfsd' process is running. %pretrans libs pidof -c -o %PPID -x glusterfsd &>/dev/null ---------> Here! if [ $? -eq 0 ]; then -------> Here! pushd . > /dev/null 2>&1 for volume in /var/lib/glusterd/vols/*; do cd $volume; vol_type=`grep '^type=' info | awk -F'=' '{print $2}'` volume_started=`grep '^status=' info | awk -F'=' '{print $2}'` if [[ $vol_type -eq 0 && $volume_started -eq 1 ]] ; then echo "ERROR: In-service rolling upgrade requires pure distribute volume(s), should be stopped" echo "ERROR: Please stop distribute volume(s) before proceeding... exiting!" exit 1; fi done popd > /dev/null 2>&1 echo "WARNING: Updating glusterfs requires its processes to be killed. This action does NOT incur downtime." echo "WARNING: Ensure to wait for the upgraded server to finish healing before proceeding." echo "WARNING: Refer install/upgrade guide for more details" echo "Please run # service glusterd stop; pkill glusterfs; pkill glusterfsd; pkill gsyncd.py;" fi Also the %pretrans by itself could be causing this issue - Niels i think this is what i was worried?
Okay so the problem here is %pretrans The issue is this - https://fedoraproject.org/wiki/Packaging:Guidelines?rd=Packaging/Guidelines#The_.25pretrans_scriptlet %pretrans is invoked very early before the dependencies are met, so a shell cannot be coded into "%pretrans" block - we should instead use "lua" http://www.rpm.org/wiki/PackagerDocs/RpmLua - i will sending out another patch.
Either that we write in "lua" or we completely avoid it by using "%pre" which basically no such restrictions. Its upto you guys to decide, i can go with either way here "lua" a new patch or removing %pretrans as we did before in RHS 2.1 Update 1
Patch is under review at https://code.engineering.redhat.com/gerrit/#/c/17611/
Please set the priority.
Patch that is supposed to fix this is part of bug 1022822.
Please set priority.
Not seeing the issue any more with glusterfs-3.4.0.51rhs-1.el6rhs. Marking it verified
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. http://rhn.redhat.com/errata/RHEA-2014-0208.html