Hi, On Wed, Dec 12, 2007 at 03:22:16PM -0500, Rolando wrote: > Hi all, > > I've had some trouble setting up Heartbeat V2 on a x86_64 Fedora Core 5 > system. I've > searched the mailing lists and linux-ha site but haven't been able to find a > solution > to my problem. > > The problem I'm having is when both nodes are active, and I try shutting down > either > to test failover. The remaining node takes over the resources, but then > immediately > stops them again. I haven't seen any obvious errors in the logs that I > haven't seen > on other working i386 nodes in other clusters.
Strange thing. Can you post both the logs and the pengine files from this period. You can grep the log to see which pengine graphs belong to this period: grep 'PEngine Input stored in' ... > > The only differences between the two versions of ha.cf are the ping and ucast > parameters > (ping = the locally attached switch, and peer = the other node). > > ha.cf: > > keepalive 1 > deadtime 10 > warntime 5 > initdead 60 > udpport 694 > ping 192.168.42.XXX # network switch IP to identify local state > ucast eth1 192.168.42.BBB # peer IP > auto_failback off > watchdog /dev/watchdog > node nfsAAA # identify each node using output of 'uname -n' > node nfsBBB # identify each node using output of 'uname -n' > use_logd yes > crm yes > apiauth cibmon uid=hacluster > respawn hacluster /usr/lib64/heartbeat/cibmon -d > > The cib.xml file was generated using the > /usr/lib64/heartbeat/haresources2cib.py script > based on the following: > > nfsAAA drbddisk::drbd-resource-0 \ > Filesystem::/dev/drbd0::/serve/spool0::ext3 \ > portmap \ > nfslock \ > nfs \ > IPaddr::10.20.30.AAA Shouldn't the IPaddr precede the network resources? > Unfortunately, the cib.xml file below does not include the secondary <node > id=..> tag, > since the systems I'm trying to set up are active production systems and I > can't afford > to have both running in a clustered mode, as this will introduce two single > points of > failure if either one goes offline. > > The only difference I was able to see between the version below and the > version generated > when both nodes were up and part of the same cluster (as verified with > crm_mon), was an > extra <node id...> tag. > > When either of the nodes is taken down to test failover, the have_quorum > attribute is "false" > and resources are shut down on the remaining online node. In a two node cluster there's always quorum. Thanks, Dejan > <cib admin_epoch="0" epoch="1" num_updates="26" have_quorum="true" > generated="true" ccm_transition="1" num_peers="1" cib_feature_revision="1.3" > dc_uuid="1dc5b2d3-7950-4ead-88f4-5fc026ce4bf3" cib-last-written="Mon Dec 10 > 02:24:40 2007"> > <configuration> > <crm_config> > <cluster_property_set id="cib-bootstrap-options"> > <attributes> > <nvpair id="cib-bootstrap-options-symmetric_cluster" > name="symmetric_cluster" value="true"/> > <nvpair id="cib-bootstrap-options-no_quorum_policy" > name="no_quorum_policy" value="stop"/> > <nvpair id="cib-bootstrap-options-default_resource_stickiness" > name="default_resource_stickiness" value="0"/> > <nvpair > id="cib-bootstrap-options-default_resource_failure_stickiness" > name="default_resource_failure_stickiness" value="0"/> > <nvpair id="cib-bootstrap-options-stonith_enabled" > name="stonith_enabled" value="false"/> > <nvpair id="cib-bootstrap-options-stonith_action" > name="stonith_action" value="reboot"/> > <nvpair id="cib-bootstrap-options-stop_orphan_resources" > name="stop_orphan_resources" value="true"/> > <nvpair id="cib-bootstrap-options-stop_orphan_actions" > name="stop_orphan_actions" value="true"/> > <nvpair id="cib-bootstrap-options-remove_after_stop" > name="remove_after_stop" value="false"/> > <nvpair id="cib-bootstrap-options-short_resource_names" > name="short_resource_names" value="true"/> > <nvpair id="cib-bootstrap-options-transition_idle_timeout" > name="transition_idle_timeout" value="5min"/> > <nvpair id="cib-bootstrap-options-default_action_timeout" > name="default_action_timeout" value="5s"/> > <nvpair id="cib-bootstrap-options-is_managed_default" > name="is_managed_default" value="true"/> > </attributes> > </cluster_property_set> > </crm_config> > <nodes> > <node id="1dc5b2d3-7950-4ead-88f4-5fc026ce4bf3" uname="nfsAAA" > type="normal"/> > </nodes> > <resources> > <group id="group_1"> > <primitive class="heartbeat" id="drbddisk_1" provider="heartbeat" > type="drbddisk"> > <operations> > <op id="drbddisk_1_mon" interval="120s" name="monitor" > timeout="60s"/> > </operations> > <instance_attributes id="drbddisk_1_inst_attr"> > <attributes> > <nvpair id="drbddisk_1_attr_1" name="1" > value="drbd-resource-0"/> > </attributes> > </instance_attributes> > </primitive> > <primitive class="ocf" id="Filesystem_2" provider="heartbeat" > type="Filesystem"> > <operations> > <op id="Filesystem_2_mon" interval="120s" name="monitor" > timeout="60s"/> > </operations> > <instance_attributes id="Filesystem_2_inst_attr"> > <attributes> > <nvpair id="Filesystem_2_attr_0" name="device" > value="/dev/drbd0"/> > <nvpair id="Filesystem_2_attr_1" name="directory" > value="/serve/spool0"/> > <nvpair id="Filesystem_2_attr_2" name="fstype" value="ext3"/> > </attributes> > </instance_attributes> > </primitive> > <primitive class="lsb" id="portmap_3" provider="heartbeat" > type="portmap"> > <operations> > <op id="portmap_3_mon" interval="120s" name="monitor" > timeout="60s"/> > </operations> > </primitive> > <primitive class="lsb" id="nfslock_4" provider="heartbeat" > type="nfslock"> > <operations> > <op id="nfslock_4_mon" interval="120s" name="monitor" > timeout="60s"/> > </operations> > </primitive> > <primitive class="lsb" id="nfs_5" provider="heartbeat" type="nfs"> > <operations> > <op id="nfs_5_mon" interval="120s" name="monitor" timeout="60s"/> > </operations> > </primitive> > <primitive class="ocf" id="IPaddr_10_20_30_AAA" provider="heartbeat" > type="IPaddr"> > <operations> > <op id="IPaddr_10_20_30_AAA_mon" interval="5s" name="monitor" > timeout="5s"/> > </operations> > <instance_attributes id="IPaddr_10_20_30_AAA_inst_attr"> > <attributes> > <nvpair id="IPaddr_10_20_30_AAA_attr_0" name="ip" > value="10.20.30.AAA"/> > </attributes> > </instance_attributes> > </primitive> > </group> > </resources> > <constraints> > <rsc_location id="rsc_location_group_1" rsc="group_1"> > <rule id="prefered_location_group_1" score="100"> > <expression attribute="#uname" id="prefered_location_group_1_expr" > operation="eq" value="nfsAAA"/> > </rule> > </rsc_location> > </constraints> > </configuration> > </cib> > > Note, the configuration generated from the haresources example above works > fine on other > i386 systems, just not on this x86_64 system. > > The heartbeat version used is heartbeat-2.0.7-3.fc5 for x86_64. > > Any help would be greatly appreciated. > > Thanks in advance, > > -Rolando > _______________________________________________ > 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
