[Pacemaker] (Linux-HA)resource-stickiness automatically move all resources in certain time

Max Williams Max.Williams at betfair.com
Wed Jun 8 16:33:03 UTC 2011


I'm not quite sure about your requirements because it sounds like you want to be able to perform a manual check of a host post failure AND have automatic fail back but perhaps you could remove the resource stickiness and look at setting migration-threshold to 1 and failure-timeout to 3600:
http://www.clusterlabs.org/doc/en-US/Pacemaker/1.1/html/Pacemaker_Explained/s-resource-options.html


From: Warrior [mailto:dareevil666 at gmail.com]
Sent: 08 June 2011 16:11
To: pacemaker at oss.clusterlabs.org
Subject: [Pacemaker] (Linux-HA)resource-stickiness automatically move all resources in certain time

Hello,
I`ve configured heartbeat and pacemaker with 2 nodes, all resources work
well, all resources start when some of them is down, when the host is down,
pacemaker moves all resources to the other one, my configuration:

node $id="06d57c5a-3d47-4ef1-b518-7b8501f5ca9d" premailman1.mpt.es<http://premailman1.mpt.es>
node $id="9af2ea3b-b7ad-458f-b7d0-2491b72db28f" premailman.mpt.es<http://premailman.mpt.es>
primitive IP-Virtual ocf:heartbeat:IPaddr \
    params ip="10.1.24.201"
primitive apache_mailman lsb:httpd \
    op monitor interval="15" timeout="15" start-delay="15" \
    meta target-role="Started"
primitive apache_webmail lsb:httpd \
    op monitor interval="15" timeout="15" start-delay="15" \
    meta target-role="Started"
primitive httpd_mailman ocf:heartbeat:anything \
    params binfile="/usr/local/bin/httpd_mailman"
primitive httpd_webmail ocf:heartbeat:anything \
    params binfile="/usr/local/bin/httpd_webmail"
primitive mailman ocf:heartbeat:anything \
    params binfile="/var/mailman/bin/mailmanctl" cmdline_options="start" \
    op start interval="0s" timeout="60s" on-fail="restart" \
    op monitor interval="10s" timeout="60s" on-fail="restart" \
    op stop interval="0s" timeout="60s" on-fail="block"
primitive smtp_postfix ocf:heartbeat:postfix \
    params binary="/etc/init.d/postfix" \
    op start interval="0s" timeout="60s" on-fail="restart" \
    op monitor interval="10s" timeout="60s" on-fail="restart" \
    op stop interval="0s" timeout="60s" on-fail="block"
colocation colocation-1 inf: IP-Virtual apache_mailman
colocation colocation-2 inf: IP-Virtual smtp_postfix
colocation colocation-3 inf: IP-Virtual mailman
colocation colocation-4 inf: IP-Virtual httpd_mailman
property $id="cib-bootstrap-options" \
    dc-version="1.1.1-b9b672590e79770afb63b9b455400d92fb6b5d9e" \
    cluster-infrastructure="Heartbeat" \
    stonith-enabled="false" \
    start-failure-is-fatal="false" \
    no-quorum-policy="ignore" \
    last-lrm-refresh="1306416533"
rsc_defaults $id="rsc-options" \
    resource-stickiness="INFINITY"

as I said everything works well, but i need just one thing, i have the
directive " resource-stickiness="INFINITY", when node1 is down, the node2
starts all resources and prevent  failback again though the node1 is up,
 back i need move all resources to the first one when I certainly sure the
node1 is runnig, is there some directive to tell pacemaker that move back
again the to node1 automatically? i mean, is posibble to move back resources
to node1 when It has passed certain time, maybe 1 hour or something?

________________________________________________________________________
In order to protect our email recipients, Betfair Group use SkyScan from 
MessageLabs to scan all Incoming and Outgoing mail for viruses.

________________________________________________________________________
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.clusterlabs.org/pipermail/pacemaker/attachments/20110608/4f4fd2f4/attachment.htm>


More information about the Pacemaker mailing list