Andrew Beekhof wrote:
Just going through the rest of my "vacation email"...
This is the same issue you filed bug 1859 for right?
OOC, what's involved in backporting this fix to heartbeat 2.1.3?
Regards,
Tim
On Mar 17, 2008, at 9:14 AM, Junko IKEDA wrote:
It's broken in 2.1.3. It was fixed in Pacemaker 0.6 though.
I run Pacemaker 0.6.2-11.1, but something is wrong with it.
My configuration is allowed to fail up to four times,
and the resource would move to the stand-by node at 5th failure.
It basically goes well, but it would run fail-over action at 3rd failure
sometime.
attached:
hb_report.tar.gz (work well)
hb_report_error.tar.gz (bad case)
node-b's "pe-input-9" in hb_report_error.tar.gz is strange.
Pengine considers the resource status as "stopped", and calculate its
score
as default status.
It loses the recorded score.
Thanks,
Junko
On Wed, Mar 12, 2008 at 4:56 PM, Jose Chavez <[EMAIL PROTECTED]>
wrote:
I've had using the command 'crm_failcount'. I guess that parameter have
to
be increasing every time the resources has a failure, but it doesn't
happen.
The first time the resource (ntp for the example) is stopped, the
parameter
fail-count is increased 0->1
Sometimes the second time the resource is stopped the parameter moves
1->2,
and sometimes doesn't change.
The next failures (3th, 4th, 5th, ......) never change the
'fail-count'.
Is this the right behaviour?
Thanks in advance.
_______________________________________________
Linux-HA mailing list
[email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems
