On Jan 9, 2008, at 12:42 AM, Ryan Pendergast wrote:
I am running a single node system, and using heartbeat (2.0.8 on
SLES) to
make sure if resources die, they will be restarted.
I am running into a problem where once a resource on my single node
cluster
has a failure count of 2, then fails again it is not restarted
(resource
sits in failed state and is no longer allowed to run on the node).
If I then reset the failure count for that resource, it instantly
starts up.
But once the failure count gets to 2, the resource stays in failed
state and
resetting failure count does not cause it to start.
I have globally set the default-resource-stickiness to INFINITY, and
have
set the default-resource-failure-stickiness to -10. In addition I
setup
rsc_location rules to make sure resources are not moved to other
nodes. I
did the above to make sure the following equation is never less than
0 (node
score + stickiness) / abs(failure stickiness). So in my case
( INFINITY +
INFINITY ) / 10 should never be < 0...
Do you know of any reason why this would be happening?
possibly a bug.
if you get the cluster into that state again (ie. before you reset the
stickiness) and attach the result of 'cibadmin -Q' i'll take a look.
(the config below isn't enough because i need the status section too)
Here are steps I do:
1) /etc/init.d/my_service stop
2) resource_mine successfully goes from failed to started state
after 1 min
3) /etc/init.d/my_service stop
4) resource_mine successfully goes from failed to started state
after 1 min
5) /etc/init.d/my_service stop
6) resource_mine goes into failed state, and cannot get out (even by
doing
crm_failcount -D -U mybox -r resource_mine)
Thanks in advance,
Ryan
Here is my cib.xml:
<cib generated="true" admin_epoch="1" epoch="2" num_updates="66"
have_quorum="true" ignore_dtd="false" num_peers="1"
cib_feature_revision="
1.3" cib-last-written="Tue Jan 8 16:54:30 2008" ccm_transition="1"
dc_uuid="768ee60e-172b-4068-bd08-d07c2176044c">
<configuration>
<crm_config>
<cluster_property_set id="cib-bootstrap-options">
<attributes>
<nvpair id="cib-bootstrap-options-last-lrm-refresh"
name="last-lrm-refresh" value="1194993736"/>
<nvpair id="cib-bootstrap-options-default-resource-
stickiness"
name="default-resource-stickiness" value="INFINITY"/>
<nvpair id="cib-bootstrap-options-default-action-timeout"
name="default-action-timeout" value="60s"/>
<nvpair
id="cib-bootstrap-options-default-resource-failure-stickiness"
name="default-resource-failure-stickiness" value="-10"/>
</attributes>
</cluster_property_set>
</crm_config>
<nodes>
<node id="768ee60e-172b-4068-bd08-d07c2176044c" uname="mybox"
type="normal"/>
</nodes>
<resources>
<group id="group_mine" ordered="false" collocated="false"
is_managed="true">
<primitive id="resource_named" class="lsb" type="named"
provider="heartbeat">
<instance_attributes id="resource_named_instance_attrs">
<attributes>
<nvpair id="resource_named_target_role"
name="target_role"
value="started"/>
</attributes>
</instance_attributes>
<operations>
<op id="named_op1" name="monitor" description="monitor
named"
interval="300" timeout="15" start_delay="300" disabled="false"
role="Started" prereq="nothing" on_fail="restart"/>
</operations>
</primitive>
<primitive id="resource_mine" class="lsb" type="my_service"
provider="heartbeat">
<instance_attributes id="resource_mine_instance_attrs">
<attributes>
<nvpair id="resource_mine_target_role"
name="target_role"
value="started"/>
</attributes>
</instance_attributes>
<operations>
<op id="my_op1" name="monitor" description="monitor my
service"
interval="60" timeout="60" disabled="false" prereq="nothing"
on_fail="restart" start_delay="60" role="Started"/>
</operations>
</primitive>
<primitive id="resource_apacheint" class="lsb"
type="apacheint"
provider="heartbeat">
<instance_attributes id="resource_apacheint_instance_attrs">
<attributes>
<nvpair id="resource_apacheint_target_role"
name="target_role" value="started"/>
</attributes>
</instance_attributes>
<operations>
<op id="apacheint_op1" name="monitor" interval="60"
timeout="30" start_delay="60" role="Started" prereq="nothing"
on_fail="restart" description="monitor internal apache"
disabled="false"/>
</operations>
</primitive>
</group>
</resources>
<constraints>
<rsc_location id="stay_1node" rsc="group_mine">
<rule id="stay_1node1" score="-INFINITY">
<expression attribute="#uname" id="stay_1node_negINF"
operation="ne" value="mybox"/>
</rule>
<rule id="stay_1node2" score="INFINITY">
<expression attribute="#uname" id="stay_1node_posINF"
operation="eq" value="mybox"/>
</rule>
</rsc_location>
</constraints>
</configuration>
</cib>
_______________________________________________
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
