Hello all,
I've been playing with heartbeat 2 for a few weeks and have a good
handle on most of the basics, but I am having trouble with some of the
subtleties. I'm trying to prevent a resource from from returning to its
preferred node until resources that are co-located with it have
SUCCESSFULLY started - it appears to be returning during the process of
starting these resources.
I have a two-node cluster with a clone filesystem resource (OCFS2) and
/two/ ip addresses. The two ips weakly prefer to run on separate nodes
on the cluster (for DNS round-robin purposes). I have set a colocation
constraint requiring that the ip resource only be allowed to run where
an instance the ocfs2 clone is (eventually this will be a web cluster).
When, say, node2 goes down, the ip address correctly fails over to
node1. However, when I bring node 2 back up OCFS often fails to mount
without rebooting (a separate issue). The problem is that the ip address
will temporarily failback to its preferred location for about 10-15
seconds while ocfs is trying to mount, then will return to its less
preferred node when ocfs actually fails.
Is there some way to prevent this behavior without setting stickiness
such that I have to manually failback the ip? I want the ip to return to
node2, but only AFTER any colocated resources have successfully started.
I tried using an order constraint, but that just had the effect of
stopping the ip resource while ocfs was attempting to mount.
Thanks in advance!
Here is my cib:
<cib generated="true" admin_epoch="0" have_quorum="true"
ignore_dtd="false" num_peers="2" cib_feature_revision="1.3"
crm_feature_set="2.0" ccm_transition="83"
dc_uuid="b2bef128-5b51-4a72-a807-48297385a8a6" epoch="383" num_updates="39">
<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:
a3184d5240c6e7032aef9cce6e5b7752ded544b3"/>
<nvpair
id="cib-bootstrap-options-default-resource-stickiness"
name="default-resource-stickiness" value="1"/>
</attributes>
</cluster_property_set>
<cluster_property_set id="cibbootstrap">
<attributes>
<nvpair id="cibbootstrap-01" name="transition_idle_timeout"
value="60"/>
<nvpair id="cibbootstrap-02"
name="default_resource_stickiness" value="1"/>
<nvpair id="cibbootstrap-03"
name="default_resource_failure_stickiness" value="-500"/>
<nvpair id="cibbootstrap-04" name="stonith_enabled"
value="false"/>
<nvpair id="cibbootstrap-05" name="stonith_action"
value="reboot"/>
<nvpair id="cibbootstrap-06" name="symmetric_cluster"
value="true"/>
<nvpair id="cibbootstrap-07" name="no_quorum_policy"
value="stop"/>
<nvpair id="cibbootstrap-08" name="stop_orphan_resources"
value="true"/>
<nvpair id="cibbootstrap-09" name="stop_orphan_actions"
value="true"/>
<nvpair id="cibbootstrap-10" name="is_managed_default"
value="true"/>
</attributes>
</cluster_property_set>
</crm_config>
<nodes>
<node id="c5583448-853e-419c-bca9-eb52e8b64c79" uname="atweb01v"
type="normal"/>
<node id="b2bef128-5b51-4a72-a807-48297385a8a6" uname="atweb02v"
type="normal"/>
</nodes>
<resources>
<clone id="ClusterStorage" notify="true" globally_unique="false">
<instance_attributes id="ClusterStorage">
<attributes>
<nvpair id="ClusterStorage-01" name="clone_node_max"
value="1"/>
<nvpair id="ClusterStorage-02" name="target_role"
value="started"/>
</attributes>
</instance_attributes>
<primitive id="OCFS2" class="ocf" type="Filesystem"
provider="heartbeat">
<operations>
<op name="monitor" interval="20s" timeout="60s"
prereq="nothing" id="OCFS2-op-01"/>
</operations>
<instance_attributes id="OCFS2">
<attributes>
<nvpair id="OCFS2-01" name="device" value="/dev/sdb"/>
<nvpair id="OCFS2-02" name="directory" value="/mnt/ocfs2"/>
<nvpair id="OCFS2-03" name="fstype" value="ocfs2"/>
<nvpair id="OCFS2:1_target_role" name="target_role"
value="started"/>
</attributes>
</instance_attributes>
</primitive>
</clone>
<primitive id="ip1" class="ocf" type="IPaddr2" provider="heartbeat">
<instance_attributes id="0f3f4fc2-354f-42c4-aca9-adfe2a4392a6">
<attributes>
<nvpair name="ip" value="140.211.89.16"
id="c68ae67b-7860-4309-a478-2c705ab2366b"/>
<nvpair name="nic" value="eth0"
id="f6db2b92-492e-442e-b084-68cc97ab7b46"/>
<nvpair name="cidr_netmask" value="24"
id="cc49f493-8e8c-48f0-bfc5-a24af7071b2c"/>
</attributes>
</instance_attributes>
<operations>
<op id="97f01f09-5322-40c1-a719-bf8b02c312fa" name="monitor"
interval="10s" timeout="20s" start_delay="5s" on_fail="restart"/>
</operations>
</primitive>
<primitive id="ip2" class="ocf" type="IPaddr2" provider="heartbeat">
<instance_attributes id="74652121-2b53-4d3d-901e-28a2d898a7bc">
<attributes>
<nvpair name="ip" value="140.211.89.17"
id="bda42825-7e28-4318-99a0-197d2f5216f0"/>
<nvpair name="nic" value="eth0"
id="55cb0faa-c05f-4a52-b1c2-fdfffb09cb12"/>
<nvpair name="cidr_netmask" value="24"
id="74420629-9086-4967-8a29-63b9dfde94a1"/>
</attributes>
</instance_attributes>
<operations>
<op id="47e5eaa3-9472-42cc-942a-f9f1d80f3443" name="monitor"
interval="10s" timeout="20s" start_delay="5s"/>
</operations>
</primitive>
</resources>
<constraints>
<rsc_colocation id="colo_ip1" from="ip1" to="ClusterStorage"
score="INFINITY"/>
<rsc_colocation id="colo_ip2" from="ip2" to="ClusterStorage"
score="INFINITY"/>
<rsc_location id="loc_ip1" rsc="ip1">
<rule id="pref_ip1" score="100">
<expression attribute="#uname" operation="eq"
value="atweb01v" id="88ce033e-90fc-4e46-959b-15d198ba0273"/>
</rule>
</rsc_location>
<rsc_location id="loc_ip2" rsc="ip2">
<rule id="pref_ip2" score="100">
<expression attribute="#uname" operation="eq"
value="atweb02v" id="876c2018-842a-41c7-8d5e-b16a66c60ea7"/>
</rule>
</rsc_location>
</constraints>
</configuration>
_______________________________________________
Linux-HA mailing list
[email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems
