Red Hat Bugzilla – Bug 615863
Improve SNMP Traps for Clustering
Last modified: 2016-04-26 11:46:11 EDT
This bug has been copied from bug #504869 and has been proposed
to be backported to 5.5 z-stream (EUS).
The initial implementation has limitations:
- there is no method to retrieve fencing information or correlate fencing events with nodes leaving the cluster; this means that there are no events generated when fencing occurs. This has been
- there is no method to retrieve service events from rgmaanger; this means that there is a chance that a trap will not be sent if a service transitions quickly. This can occur, for example, if a service fails and is restarted prior to the SNMP module updating its state
- rgmanager has no method to report when individual resources fail.
- the way libclustersnmp gains cluster node information is by polling; this means that there is a chance that a trap will not be sent if a node transitions very quickly (e.g. service cman restart).
Development Management has reviewed and declined this request. You may appeal
this decision by reopening this request.