If both servers involved in a failed process are managed
by IBM® Sterling Control Center
Monitor, two alerts may be generated.
About this task
The redundant alert can be eliminated by setting up a rule
that causes IBM Sterling Control Center
Monitor to ignore one of the two alert conditions.
Procedure
- Create a rule that specifies the following parameters:
| Key |
Operator |
Value |
| Event Type |
Matches |
Process Step Ended |
| Return Code |
Not equal to |
0 |
| Local Node (P or S) |
Matches |
S |
- Specify an action of “No operation” and finish by confirming
your choices.
Attention: Except for the additional parameter of “Local Node (P or S) Matches S,” this
rule is essentially a duplicate of the Bad Return Code rule that ships with IBM Sterling Control Center
Monitor. To work,
it must be placed at the top of the Bad Return Code rule in the Rules listing, In other words, it
must a have a higher priority in the listing.
CAUTION:
This technique
works except in the following situations:
- If a PNODE = SNODE Process step fails, two alerts are still generated.
- No alerts are generated if the PNODE server involved in the Process
is unmanaged. To solve this problem, move a rule that specifies a
particular server as the SNODE for a Process before this rule in the
Rules listing.