Any chance you could attach the cib from just before you set is-managed-default to true?
On Tue, Nov 4, 2008 at 01:49, Alex Strachan <[EMAIL PROTECTED]> wrote: > HA 2.99.1, pacemaker 1.0, RHES4 64bit. > > My basic HA setup consists of one resource group and two nodes. > > When performing maintenance on the system our approach in the past has been > to follow this procedure. > Set all resources to unmanaged via custom script - ha_freeze.sh > Manually control the resources > Clear any fail counts which have been recorded against the resources > Set all resources to managed via custom script - ha_thaw.sh > > Previously I was able to > Set all resources to unmanaged > Clear any fail-counts > Set all resources to managed > and the cluster didn't take any actions, now with this version of the > software the cluster stop/start most of the resources consistently. > > Are others finding the same behavior? > > > Example > > [EMAIL PROTECTED] hb]# crm_resource -L -V > crm_resource[6523]: 2008/11/04_10:27:40 notice: unpack_config: On loss of > CCM Quorum: Ignore > Resource Group: group_its > resource_its_drbd (heartbeat:its_drbddisk) Started > resource_its_fs (ocf::heartbeat:its_Filesystem) Started > resource_its_vip (ocf::heartbeat:IPaddr) Started > resource_its_oracle (ocf::heartbeat:its_oracle) Started > resource_its_oralsnr (ocf::heartbeat:its_oralsnr) Started > resource_its_aims (lsb:its_aims) Started > resource_its_apache (ocf::heartbeat:its_apache) Started > resource_its_smb (lsb:its_smb) Started > resource_its_dhcpd (lsb:its_dhcpd) Started > > Set all resources to unmanaged ... > for myr in resource_its_dhcpd resource_its_smb resource_its_apache > resource_its_aims resource_its_oralsnr resource_its_oracle resource_its_vip > resource_its_fs resource_its_drbd; do crm_resource -p is-managed -r $myr > -t primitive -v false; sleep 10; done > > Now we have ... > Resource Group: group_its > resource_its_drbd (heartbeat:its_drbddisk): Started dtbaims > (unmanaged) > resource_its_fs (ocf::heartbeat:its_Filesystem): Started > dtbaims (unmanaged) > resource_its_vip (ocf::heartbeat:IPaddr): Started dtbaims > (unmanaged) > resource_its_oracle (ocf::heartbeat:its_oracle): Started dtbaims > (unmanaged) > resource_its_oralsnr (ocf::heartbeat:its_oralsnr): Started > dtbaims (unmanaged) > resource_its_aims (lsb:its_aims): Started dtbaims (unmanaged) > resource_its_apache (ocf::heartbeat:its_apache): Started dtbaims > (unmanaged) > resource_its_smb (lsb:its_smb): Started dtbaims (unmanaged) > resource_its_dhcpd (lsb:its_dhcpd): Started dtbaims (unmanaged) > > Looks good > > Check for fail-counts > [EMAIL PROTECTED] hb]# cibadmin -Q | grep 'name="fail-count' > > Clear - set all resources to managed. My expectation is that there will be > no HA changes. > > [EMAIL PROTECTED] hb]# for myr in resource_its_dhcpd resource_its_smb > resource_its_apache resource_its_aims resource_its_oralsnr > resource_its_oracle resource_its_vip resource_its_fs resource_its_drbd; do > crm_resource -p is-managed -r $myr -t primitive -v true; sleep 10; done > > The HA does re-act, it stops resources down to its_vip, then restarts. > > [EMAIL PROTECTED] hb]# crm_resource -L -V > crm_resource[6605]: 2008/11/04_10:35:07 notice: unpack_config: On loss of > CCM Quorum: Ignore > Resource Group: group_its > resource_its_drbd (heartbeat:its_drbddisk) Started > resource_its_fs (ocf::heartbeat:its_Filesystem) Started > resource_its_vip (ocf::heartbeat:IPaddr) Started > resource_its_oracle (ocf::heartbeat:its_oracle) Started > resource_its_oralsnr (ocf::heartbeat:its_oralsnr) Started > resource_its_aims (lsb:its_aims) Started > resource_its_apache (ocf::heartbeat:its_apache) Stopped > resource_its_smb (lsb:its_smb) Stopped > resource_its_dhcpd (lsb:its_dhcpd) Stopped > > > During these operations I get > Nov 4 10:30:06 itbaims crmd: [7779]: ERROR: te_graph_trigger: Transition > failed: terminated > ... > in /var/log/messages. > > > I have attached the logs from /var/log/messages, cib.xml and custom scripts. > > -- > Alex > > > _______________________________________________ > Linux-HA mailing list > [email protected] > http://lists.linux-ha.org/mailman/listinfo/linux-ha > See also: http://linux-ha.org/ReportingProblems > _______________________________________________ Linux-HA mailing list [email protected] http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
