Sorry for the delay...
On Dec 17, 2007, at 12:08 PM, Junko IKEDA wrote:
Hi,
I am testing some monitor operation behaviors,
and find a different one between 2.1.2 and dev(2.1.3).
when I run 2.1.2, it incremented a fail count if a monitor operation
timed
out.
but with dev, it doesn't up the count.
Is it an expected change?
Originally I thought so, but I believe you are correct and that the
failcount should be incremented.
The "problem" was that operations that time out now have rc = -2...
which was triggering some logic that had been trying to filter out
pending operations (which is now done differently).
I believe this is the correct patch
diff -r c3990dd76fca transitioner/events.c
--- a/transitioner/events.c Mon Dec 17 10:11:56 2007 +0100
+++ b/transitioner/events.c Tue Dec 18 14:18:52 2007 +0100
@@ -532,7 +532,7 @@ process_graph_event(crm_data_t *event, c
crm_debug_2("Processed update to %s: %s", id, magic);
}
- if(passed == FALSE && rc > EXECRA_OK) {
+ if(passed == FALSE && rc != EXECRA_OK) {
update_failcount(event, event_node, rc);
}
Monitor NG, like removing its statement file, would up the fail count.
Monitor timeout should be the same.
Best Regards,
Junko Ikeda
NTT DATA INTELLILINK CORPORATION
<
hb_report
-2.1.2
.tar
.gz
>
<
hb_report
-2.1.3
.tar.gz><Dummy.diff>_______________________________________________
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
