On a fedora 19 system where I had 3.4.4 all-in-one, I was able to upgrade to 3.5.
vdsm part had problems I was able to solve reinstalling host from the webadmin gui
Steps were:
- follow instructions for engine part --> ok
- verify connectivity of engine via new webadmin
- put host into maintenance
- stop vdsmd service
- yum update vdsm part and all other packages related to ovirt
When trying to start vdsm again I got
[root@tekkaman vdsm]# systemctl status vdsmd
vdsmd.service - Virtual Desktop Server Manager
Loaded: loaded (/usr/lib/systemd/system/vdsmd.service; enabled)
Active: failed (Result: start-limit) since Sat 2014-10-25 16:01:29 CEST; 2min 54s ago
Process: 14448 ExecStartPre=/usr/libexec/vdsm/vdsmd_init_common.sh --pre-start (code=exited, status=1/FAILURE)
Oct 25 16:01:29 tekkaman.localdomain.local systemd[1]: Failed to start Virtual Desktop Server Manager.
Oct 25 16:01:29 tekkaman.localdomain.local systemd[1]: Unit vdsmd.service entered failed state.
Oct 25 16:01:29 tekkaman.localdomain.local systemd[1]: vdsmd.service holdoff time over, scheduling restart.
Oct 25 16:01:29 tekkaman.localdomain.local systemd[1]: Stopping Virtual Desktop Server Manager...
Oct 25 16:01:29 tekkaman.localdomain.local systemd[1]: Starting Virtual Desktop Server Manager...
Oct 25 16:01:29 tekkaman.localdomain.local systemd[1]: vdsmd.service start request repeated too quickly, refusing to start.
Oct 25 16:01:29 tekkaman.localdomain.local systemd[1]: Failed to start Virtual Desktop Server Manager.
Oct 25 16:01:29 tekkaman.localdomain.local systemd[1]: Unit vdsmd.service entered failed state.
I was not able to check wich part of --pre-start functions had failed..
I see inside the script
case "$1" in
--pre-start)
run_tasks " \
mkdirs \
configure_coredump \
configure_vdsm_logs \
run_init_hooks \
check_is_configured \
validate_configuration \
prepare_transient_repository \
syslog_available \
nwfilter \
dummybr \
load_needed_modules \
tune_system \
test_space \
test_lo \
unified_network_persistence_upgrade \
restore_nets \
upgrade_300_nets \
"
;;
In logs I had only supervdsm.log
MainThread::DEBUG::2014-10-25 16:34:34,021::supervdsmServer::451::SuperVdsm.Server::(main) Terminated normally
MainThread::DEBUG::2014-10-25 16:35:21,836::netconfpersistence::134::root::(_getConfigs) Non-existing config set.
MainThread::DEBUG::2014-10-25 16:35:21,836::netconfpersistence::134::root::(_getConfigs) Non-existing config set.
MainThread::DEBUG::2014-10-25 16:35:21,898::supervdsmServer::411::SuperVdsm.Server::(main) Making sure I'm root - SuperVdsm
MainThread::DEBUG::2014-10-25 16:35:21,899::supervdsmServer::420::SuperVdsm.Server::(main) Parsing cmd args
MainThread::DEBUG::2014-10-25 16:35:21,899::supervdsmServer::423::SuperVdsm.Server::(main) Cleaning old socket /var/run/vdsm/svdsm.sock
MainThread::DEBUG::2014-10-25 16:35:21,899::supervdsmServer::427::SuperVdsm.Server::(main) Setting up keep alive thread
MainThread::DEBUG::2014-10-25 16:35:21,899::supervdsmServer::433::SuperVdsm.Server::(main) Creating remote object manager
MainThread::DEBUG::2014-10-25 16:35:21,901::fileUtils::192::Storage.fileUtils::(chown) Changing owner for /var/run/vdsm/svdsm.sock, to (36:36)
MainThread::DEBUG::2014-10-25 16:35:21,901::supervdsmServer::444::SuperVdsm.Server::(main) Started serving super vdsm object
sourceRoute::DEBUG::2014-10-25 16:35:21,901::sourceroutethread::75::root::(_subscribeToInotifyLoop) sourceRouteThread.subscribeToInotifyLoop started
net_persistence = ifcfg
even after reboot only supervdsm.log was populated...
Nevertheless a host reinstall from webadmin gui seemed to fix the problem and after that it was able to start vdsm ad host part.
I was also able to upgrade DC/Cluster from 3.4 to 3.5 version.
Now that I'm on 3.5 next days I'll try to pass from F19 to F20... we'll see
Gianluca