Hi, On Mon, Jun 15, 2009 at 08:25:14PM +0530, Lasantha Marian wrote: > Hi, > > I have a two node cluster setup using Heartbeat 2.99.3/Pacemaker 1.0.2 > on an Ubuntu 8.10 server built on two identical Dell 2900 servers. > Basically Heartbeat works as expected. My configuration uses CRM to > manage and DRBD for storages. > > I have noticed the following. > > If node A is running and when node node B has failed, and then if node A > is rebooted (whilst node B still reaming failed), all the resources on > the node A would not start. Is this the expected behavior ? But I was > expecting node A to bring all resources up even when node B is being > temporarily unavailable. Can this be achieved ? > > If both nodes (A and B) are started again, all resources come up correctly. > > In the log it mostly shows "WARN: native_color: Resource dr0:1 cannot > run anywhere" and similar.
You don't have fencing (stonith) configured. With shared storage it's a must. It's a must anyway with only two nodes. Though you can disable stonith in properties, while you're testing and don't care about the data. Thanks, Dejan > Any help/explanation is appreciated. > > Best regards, > > Lasantha. > primitive dr0 ocf:heartbeat:drbd \ > params drbd_resource="www" \ > op monitor interval="59s" role="Master" timeout="30s" \ > op monitor interval="60s" role="Slave" timeout="30s" > primitive dr1 ocf:heartbeat:drbd \ > params drbd_resource="spool" \ > op monitor interval="59s" role="Master" timeout="30s" \ > op monitor interval="60s" role="Slave" timeout="30s" > primitive dr3 ocf:heartbeat:drbd \ > params drbd_resource="data" \ > op monitor interval="59s" role="Master" timeout="30s" \ > op monitor interval="60s" role="Slave" timeout="30s" > primitive fs0 ocf:heartbeat:Filesystem \ > params fstype="ext3" directory="/ha/www" device="/dev/drbd0" > options="rw,relatime" > primitive fs1 ocf:heartbeat:Filesystem \ > params fstype="ext3" directory="/ha/spool" device="/dev/drbd1" > options="rw,relatime" > primitive fs3 ocf:heartbeat:Filesystem \ > params fstype="ext3" directory="/ha/data" device="/dev/drbd3" > options="rw,relatime" > primitive ip0 ocf:heartbeat:IPaddr2 \ > params ip="192.168.0.180" broadcast="192.168.0.255" nic="bond0" > iflabel="0" cidr_netmask="24" \ > op monitor interval="21s" timeout="5s" > primitive pg0 ocf:heartbeat:pgsql \ > params pgctl="/usr/local/PostgreSQL/bin/pg_ctl" > psql="/usr/local/PostgreSQL/bin/psql" pgdata="/ha/data/PostgreSQL" > pgdba="postgresql" pgport="5432" pgdb="template1" \ > op monitor interval="10s" timeout="10s" > group rg0 fs0 fs1 fs3 ip0 pg0 > ms ms0 dr0 \ > meta clone-max="2" notify="true" globally-unique="false" > target-role="Started" > ms ms1 dr1 \ > meta clone-max="2" notify="true" globally-unique="false" > target-role="Started" > ms ms3 dr3 \ > meta clone-max="2" notify="true" globally-unique="false" > target-role="Started" > location ms0-master-on-connected-node ms0 \ > rule $id="ms0-master-on-connected-node-rule" $role="master" -inf: > not_defined pingd or pingd lte 0 > location ms1-master-on-connected-node ms1 \ > rule $id="ms1-master-on-connected-node-rule" $role="master" -inf: > not_defined pingd or pingd lte 0 > location ms3-master-on-connected-node ms3 \ > rule $id="ms3-master-on-connected-node-rule" $role="master" -inf: > not_defined pingd or pingd lte 0 > colocation rg0-on-ms0 inf: rg0 ms0:Master > colocation rg0-on-ms1 inf: rg0 ms1:Master > colocation rg0-on-ms3 inf: rg0 ms3:Master > order ms0-before-rg0 inf: ms0:promote rg0:start > order ms1-before-rg0 inf: ms1:promote rg0:start > order ms3-before-rg0 inf: ms3:promote rg0:start > property $id="cib-bootstrap-options" \ > dc-version="1.0.2-c02b459053bfa44d509a2a0e0247b291d93662b7" \ > no-quorum-policy="ignore" > _______________________________________________ > 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
