In addition, 'default-resource-stickiness' is set to INFINITY in my cib.xml .
# grep stickiness /var/lib/heartbeat/crm/cib.xml
<nvpair name="default-resource-stickiness"
id="cib-bootstrap-options-default-resource-stickiness"
value="INFINITY"/>
<nvpair name="default-resource-failure-stickiness"
id="cib-bootstrap-options-default-resource-failure-stickiness"
value="0"/>
As I read the documentation at
http://linux-ha.org/v2/dtd1.0/annotated#default_resource_stickiness,
'INFINITY' means that the resource should remain at their current
location, not flip back to the standby host.
-= Stefan
On 7/12/07, Stefan Lasiewski <[EMAIL PROTECTED]> wrote:
I'm using heartbeat heartbeat-2.0.8-2.el4.centos on two Redhat 4u4 systems.
I can fail services from the 'primary' server to the 'standby' server.
I can also fail services back from the 'standby' to the 'primary'.
However, if I restart the 'standby', then the 'standby' takes over all
of the resources again.
Here is my scenario:
- Server fs1 is the primary server.
- Server fs2 is the standby server.
- According to 'crm_mon', both servers are 'online'. If 'fs2' is 'standby',
then failover never occurs.
Node: fs2 (9ca53dad-c180-4f8c-ba70-a2b0becc7677): online
Node: fs1 (ebfa769d-cab6-497b-9b82-23dc1d150ead): online
- To simulate a server failure, I power off 'fs1'.
- The resources properly fail over to fs2.
Now I want to fail back to fs1.
The following sequence of commands seems to force all service back onto 'fs1'.
# crm_standby --node-uname fs2 --attr-value on
This works great. fs1 becomes primary for all resources.
However, if I restart the machine 'fs2' or run '/etc/init.d/heartbeat
restart' on fs2, then fs2 takes over all resources again. This only
happens some of the time, not all of the time.
What am I missing?
-= Stefan
--
Stefan Lasiewski
http://stefanco.com
_______________________________________________
Linux-HA mailing list
[email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems
