Lars, Did you go to the links ?
http://linux-ha.org/FAQ#heavy_load - Leads to a page with a watermark "This web page is no longer maintained. The information presented here exists only to avoid breaking historical links" Matthew Soffen Lead Computer Scientist ISO New England - http://www.iso-ne.com/ -----Original Message----- From: [email protected] [mailto:[email protected]] On Behalf Of Lars Ellenberg Sent: Thursday, September 08, 2011 12:10 PM To: [email protected] Subject: Re: [Linux-HA] Primary node restart on Standby node link failure? On Wed, Sep 07, 2011 at 08:54:36AM -0700, Michael Moon wrote: > I searched hi and low and couldn't find an answer to this issue. We > are running heartbeat 2.1.4 on a two node configuration. If I stop the > both ethernet interfaces on the Standby node, the Application on the > Primary node will restart. Anyone know why this would be happening? I > would think there would be a setting on what to do upon a failure, > (ie, restart, nothing, etc...) but I can't find it for this version of > heartbeat. In the log you posted below, there are two urls. Could you be bothered to visit them? Have you? If they unclear, could you point out where, and how can we improve the wording so it will be clear? Anyways, if your intention was to test node failure, you should test node failure, not total cluster communication breakdown aka split-brain, which is what you do here. If your intention actually was to test split-brain, you should be aware of the implications, and what to expect. The only remotely useful thing heartbeat in haresources mode can do to recover from a split brain is to stop everything, everywhere, and restart from scratch. Which is what you observe here. crm mode, aka Pacemaker won't handle this so much differently: it will restart all resources and dependencies that seem to run more often than they should, according to the then merge configuration. To avoid split brain, you need to implement fencing, typically that means stonith. You should also read http://ourobengr.com/ha > Thanks in advance. > > > The ha.cf looks like > > logfile /var/log/ha-log > logfacility local0 > ucast eth0 10.xxx.xx.16 > ucast eth0 10.xx.xx.17 > ucast eth1 169.254.10.1 > ucast eth1 169.254.10.2 > auto_failback off > node customer-site-node1 > node customer-site-node2 > deadtime 60 > keepalive 1 > warntime 5 > > > Here is the ha-log > > heartbeat[13008]: 2011/09/06_17:41:30 info: Link customer-site-node2:eth0 > dead. > heartbeat[13008]: 2011/09/06_17:41:31 WARN: node customer-site-node2: is dead > heartbeat[13008]: 2011/09/06_17:41:31 WARN: No STONITH device configured. > heartbeat[13008]: 2011/09/06_17:41:31 WARN: Shared disks are not protected. > heartbeat[13008]: 2011/09/06_17:41:31 info: Resources being acquired from > customer-site-node2. > heartbeat[13008]: 2011/09/06_17:41:31 info: Link customer-site-node2:eth1 > dead. > harc[20623]: 2011/09/06_17:41:31 info: Running /etc/ha.d/rc.d/status status > mach_down[20653]: 2011/09/06_17:41:31 info: > /usr/share/heartbeat/mach_down: nice_failback: foreign resources acqui > red > mach_down[20653]: 2011/09/06_17:41:31 info: mach_down takeover complete > for node customer-site-node2. > heartbeat[13008]: 2011/09/06_17:41:31 info: mach_down takeover complete. > IPaddr[20697]: 2011/09/06_17:41:31 INFO: Running OK > heartbeat[20624]: 2011/09/06_17:41:31 info: Local Resource acquisition > completed. > heartbeat[13008]: 2011/09/06_17:41:47 CRIT: Cluster node customer-site-node2 > returning after partition. > heartbeat[13008]: 2011/09/06_17:41:47 info: For information on cluster > partitions, See URL: http://linux-ha.org/SplitBra > in > heartbeat[13008]: 2011/09/06_17:41:47 WARN: Deadtime value may be too small. > heartbeat[13008]: 2011/09/06_17:41:47 info: See FAQ for information on tuning > deadtime. > heartbeat[13008]: 2011/09/06_17:41:47 info: URL: > http://linux-ha.org/FAQ#heavy_load > heartbeat[13008]: 2011/09/06_17:41:47 info: Link customer-site-node2:eth0 up. > heartbeat[13008]: 2011/09/06_17:41:47 WARN: Late heartbeat: Node > customer-site-node2: interval 76500 ms > heartbeat[13008]: 2011/09/06_17:41:47 info: Status update for node > customer-site-node2: status active > harc[20765]: 2011/09/06_17:41:47 info: Running /etc/ha.d/rc.d/status status > heartbeat[13008]: 2011/09/06_17:41:49 info: Link customer-site-node2:eth1 up. > heartbeat[13008]: 2011/09/06_17:41:49 WARN: Shutdown delayed until current > resource activity finishes. > heartbeat[13008]: 2011/09/06_17:41:50 info: Heartbeat shutdown in progress. > (13008) > heartbeat[13008]: 2011/09/06_17:41:50 info: Received shutdown notice from > 'customer-site-node2'. > heartbeat[13008]: 2011/09/06_17:41:50 info: Resource takeover cancelled - > shutdown in progress. > _______________________________________________ > Linux-HA mailing list > [email protected] > http://lists.linux-ha.org/mailman/listinfo/linux-ha > See also: http://linux-ha.org/ReportingProblems -- : Lars Ellenberg : LINBIT | Your Way to High Availability : DRBD/HA support and consulting http://www.linbit.com DRBD® and LINBIT® are registered trademarks of LINBIT, Austria. _______________________________________________ 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
