[Pacemaker] on fail standby with migration-threshold

rvm at free.fr rvm at free.fr
Tue Feb 16 06:41:58 EST 2010


Hi all,

I'm trying to have this configuration on my 2 nodes test cluster :
* After two failures on a resource, standby the node and move the resource away. In fact this :

primitive APACHE ocf:me:apache \
	params url="http://localhost/server-status" \
	op monitor interval="10s" timeout="30s" \
	meta target-role="Started" migration-threshold="2" on-fail="standby"

With this conf, the node goes directly in standby mode...

or if it's impossible :
* After two failures on a resource, move the resource away and _never_ go back even if the new node fails.
=> colocation ?

Any suggestions ?

Thanks you,
Regards


My test cluster :
Debian Lenny with :
* heartbeat 3.0.beta1+hg20091102-2~bpo50+1
* pacemaker 1.0.6+hg20091102-4~bpo50+1

And my full conf :
crm(live)configure# show
node $id="1894b6a0-6648-4426-91d3-863baf0edf33" ha-test1 \
	attributes standby="off"
node $id="abb9493e-6c82-4b20-a417-ab23089275b9" ha-test2 \
	attributes standby="off"
primitive APACHE ocf:me:apache \
	params url="http://localhost/server-status" \
	op monitor interval="10s" timeout="30s" on-fail="standby" \
	meta target-role="Started" migration-threshold="2"
property $id="cib-bootstrap-options" \
	stonith-enabled="false" \
	no-quorum-policy="ignore" \
	default-resource-stickiness="100" \
	dc-version="1.0.6-cebe2b6ff49b36b29a3bd7ada1c4701c7470febe" \
	cluster-infrastructure="Heartbeat" \
	last-lrm-refresh="1266319346"



More information about the Pacemaker mailing list