Michael Judd wrote:
> Hi,
>
> I'm using the ocf/heartbeat resource (Master/Slave OCF Resource Agent
> for DRBD) with a ocf/heartbeat Filesystem resource to mount the
> filesystem which is ext3.
>
> This can be mounted and unmounted on each node without hearbeat.
>
> However, when I fail over a node, the new (acquiring) node appears to be
> unable to mount the node with a message as follows:
>
> crmd[2184]: 2009/03/27_14:08:02 info: do_lrm_rsc_op: Performing
> op=FS_drbd0_start_0 key=41:15:2b5cddee-44fb-4857-93ff-69172171e2da)
> lrmd[2181]: 2009/03/27_14:08:02 info: rsc:FS_drbd0: start
> lrmd[2181]: 2009/03/27_14:08:03 info: RA output:
> (DRBD_drbd0:1:notify:stdout) No set matching
> id=master-9ed2c6ec-f548-4505-b332-e2e715083ac0 in status
>
> crmd[2184]: 2009/03/27_14:08:03 info: process_lrm_event: LRM operation
> DRBD_drbd0:1_notify_0 (call=19, rc=0) complete
> Filesystem[4442]: 2009/03/27_14:08:03 INFO: Running start for /dev/drbd0
> on /var/lib/pgsql
> lrmd[2181]: 2009/03/27_14:08:03 info: RA output: (FS_drbd0:start:stderr)
> /dev/drbd0: Wrong medium type
>
> lrmd[2181]: 2009/03/27_14:08:03 info: RA output: (FS_drbd0:start:stderr)
> mount: block device /dev/drbd0 is write-protected, mounting read-only
>
> lrmd[2181]: 2009/03/27_14:08:03 info: RA output: (FS_drbd0:start:stderr)
> mount: Wrong medium type
>
> Filesystem[4442]: 2009/03/27_14:08:04 ERROR: Couldn't mount filesystem
> /dev/drbd0 on /var/lib/pgsql
> crmd[2184]: 2009/03/27_14:08:04 ERROR: process_lrm_event: LRM operation
> FS_drbd0_start_0 (call=21, rc=1) Error unknown error
> crmd[2184]: 2009/03/27_14:08:04 info: do_lrm_rsc_op: Performing
> op=DRBD_drbd0:1_notify_0 key=54:15:2b5cddee-44fb-4857-93ff-69172171e2da)
> lrmd[2181]: 2009/03/27_14:08:04 info: rsc:DRBD_drbd0:1: notify
>
> If I perform a cleanup on the filesystem resource, then the service
> comes up correctly. It appears the drbd resource may not demoting the
> old node correctly. When mounting the device manually, I received an
> error message saying that both nodes could not be primary.
>
> Any ideas on how to solve this?
>
> /etc/drbd.conf
> global {
> usage-count yes;
> }
>
> resource drbd0 {
> protocol C;
>
> startup { wfc-timeout 0; # infinitely wait on resources during init
> degr-wfc-timeout 120; }
> disk { on-io-error detach; }
> net { timeout 60; connect-int 10; ping-int 10;
> max-buffers 2048; # data block buffers used before writing
> to disk
> max-epoch-size 2048;
> shared-secret "C45hgtreeu!3z";
> ko-count 4; # Peer is dead is this count is exceeded
> }
> syncer { rate 40M; } # sync
>
> on pgcluster1.weir.lan {
> device /dev/drbd0;
> disk /dev/hdb1;
> address 192.168.1.101:7788;
> meta-disk internal;
> }
> on pgcluster2.weir.lan {
> device /dev/drbd0;
> disk /dev/hdb1;
> address 192.168.1.102:7788;
> meta-disk internal;
> }
>
> /var/lib/heartbeat/crm/cib.xml
>
> <cib generated="true" admin_epoch="0" have_quorum="true"
> ignore_dtd="false" num_peers="2" cib_feature_revision="2.0"
> ccm_transition="34" dc_uuid="9ed2c6ec-f548-4505-b332-e2e715083ac0"
> epoch="161" num_updates="1" cib-last-written="Fri Mar 27 14:11:10 2009">
> <configuration>
> <crm_config>
> <cluster_property_set id="cib-bootstrap-options">
> <attributes>
> <nvpair id="cib-bootstrap-options-dc-version"
> name="dc-version" value="2.1.3-node:
> 552305612591183b1628baa5bc6e903e0f1e26a3"/>
Please upgrade. 2.1.3 has a lot of bugs in Master/Slave management.
> <nvpair name="last-lrm-refresh"
> id="cib-bootstrap-options-last-lrm-refresh" value="1238163065"/>
> </attributes>
> </cluster_property_set>
> </crm_config>
> <nodes>
> <node uname="pgcluster2.weir.lan" type="normal"
> id="9ed2c6ec-f548-4505-b332-e2e715083ac0">
> <instance_attributes
> id="nodes-9ed2c6ec-f548-4505-b332-e2e715083ac0">
> <attributes>
> <nvpair name="standby"
> id="standby-9ed2c6ec-f548-4505-b332-e2e715083ac0" value="off"/>
> </attributes>
> </instance_attributes>
> </node>
> <node id="a28b1a3d-b34b-4791-b91c-ad77c7ff428c"
> uname="pgcluster1.weir.lan" type="normal">
> <instance_attributes
> id="nodes-a28b1a3d-b34b-4791-b91c-ad77c7ff428c">
> <attributes>
> <nvpair id="standby-a28b1a3d-b34b-4791-b91c-ad77c7ff428c"
> name="standby" value="off"/>
> </attributes>
> </instance_attributes>
> </node>
> </nodes>
> <resources>
> <master_slave id="ms-drbd0">
> <meta_attributes id="ms-drbd0_meta_attrs">
> <attributes>
> <nvpair id="ms-drbd0_metaattr_target_role"
> name="target_role" value="started"/>
> <nvpair id="ms-drbd0_metaattr_clone_max" name="clone_max"
> value="2"/>
> <nvpair id="ms-drbd0_metaattr_clone_node_max"
> name="clone_node_max" value="1"/>
> <nvpair id="ms-drbd0_metaattr_master_max" name="master_max"
> value="1"/>
> <nvpair id="ms-drbd0_metaattr_master_node_max"
> name="master_node_max" value="1"/>
> <nvpair id="ms-drbd0_metaattr_notify" name="notify"
> value="true"/>
> <nvpair id="ms-drbd0_metaattr_globally_unique"
> name="globally_unique" value="false"/>
> </attributes>
> </meta_attributes>
> <primitive id="DRBD_drbd0" class="ocf" type="drbd"
> provider="heartbeat">
> <instance_attributes id="DRBD_drbd0_instance_attrs">
> <attributes>
> <nvpair id="f0deba6f-9513-4e34-9f8d-b949fe898b2b"
> name="drbd_resource" value="drbd0"/>
> </attributes>
> </instance_attributes>
> <operations>
> <op id="0ec050a8-2fc0-4f91-8ac6-52e8f12fd341"
> name="monitor" description="op-drbd0-1" interval="59s" timeout="10s"
> start_delay="1m" role="Master" disabled="false"/>
> <op id="2a7499c3-5078-41ce-bff2-afdfae78a584"
> name="monitor" description="op-drbd0-2" interval="60s" timeout="10s"
> start_delay="1m" role="Slave" disabled="false"/>
> </operations>
> <meta_attributes id="DRBD_drbd0:1_meta_attrs">
> <attributes>
> <nvpair id="DRBD_drbd0:0_metaattr_target_role"
> name="target_role" value="started"/>
> </attributes>
> </meta_attributes>
> </primitive>
> </master_slave>
> <group id="PG_Group">
> <meta_attributes id="PG_Group_meta_attrs">
> <attributes>
> <nvpair name="target_role"
> id="PG_Group_metaattr_target_role" value="started"/>
> <nvpair id="PG_Group_metaattr_ordered" name="ordered"
> value="true"/>
> <nvpair id="PG_Group_metaattr_collocated" name="collocated"
> value="true"/>
> </attributes>
> </meta_attributes>
> <primitive id="IP_100" class="ocf" type="IPaddr2"
> provider="heartbeat">
> <instance_attributes id="IP_100_instance_attrs">
> <attributes>
> <nvpair id="9e6fbfde-bc59-48ff-b188-28c8f66f6bdd"
> name="ip" value="192.168.1.100"/>
> <nvpair id="0a3c6f65-0572-480e-9d27-16c8be4ec018"
> name="nic" value="eth0:1"/>
> </attributes>
> </instance_attributes>
> <meta_attributes id="IP_100_meta_attrs">
> <attributes>
> <nvpair id="IP_100_metaattr_target_role"
> name="target_role" value="started"/>
> </attributes>
> </meta_attributes>
> </primitive>
> <primitive id="FS_drbd0" class="ocf" type="Filesystem"
> provider="heartbeat">
> <instance_attributes id="FS_drbd0_instance_attrs">
> <attributes>
> <nvpair id="a13a908f-25b9-4e49-9446-368d4485ac59"
> name="device" value="/dev/drbd0"/>
> <nvpair id="da3c0716-6678-48ce-a544-41dda7e8734e"
> name="directory" value="/var/lib/pgsql"/>
> <nvpair id="95b5966c-cc3d-42e4-a884-3c6dec64c3ec"
> name="fstype" value="ext3"/>
> </attributes>
> </instance_attributes>
> <meta_attributes id="FS_drbd0_meta_attrs">
> <attributes>
> <nvpair id="FS_drbd0_metaattr_target_role"
> name="target_role" value="started"/>
> </attributes>
> </meta_attributes>
> </primitive>
> <primitive id="PG_service" class="lsb" type="postgresql"
> provider="heartbeat">
> <meta_attributes id="PG_service_meta_attrs">
> <attributes>
> <nvpair id="PG_service_metaattr_target_role"
> name="target_role" value="started"/>
> </attributes>
> </meta_attributes>
> </primitive>
> </group>
> </resources>
> <constraints>
> <rsc_location id="PG_location" rsc="PG_Group">
> <rule id="prefered_PG_location" score="100">
> <expression attribute="#uname"
> id="3f381e7f-9d78-4744-b872-53974241612e" operation="eq"
> value="pgcluster1.weir.lan"/>
> </rule>
> </rsc_location>
> </constraints>
You're also lacking the necessary order and colocation constraints.
Please upgrade your cluster software and then refer to
http://www.clusterlabs.org/wiki/DRBD_HowTo_1.0 for configuration.
Regards
Dominik
> </configuration>
> </cib>
>
>
>
_______________________________________________
Linux-HA mailing list
[email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems
