If you have a problem with a newer client on an older instance you would get some ora-errors.
I am not sure if all sections could run async.
Is it still the case that any ORA* service changes to UNKN sporadicly? Can you provide the oracle section of your agent output at such a situation? Please check the output for sensitive data before posting it.
I think I now understood the problem but still don’t get it how to monitor the RAC. As per this,
I am monitoring my Oracle RAC via the SCAN listener whereas the documentation says not to do that.
Since, I am monitoring it via a 3rd server, I don’t know what else to configure in mk_oracle.cfg. Is there a special setting for Oracle RAC for remote instance monitoring ?
I still feel that its not possible to achieve using the 3rd server. Right ?
The problem with RAC and SCAN listener is, that every check period you could potentially end on a different node.
Beside this problem you have some checks which have to be done on each node/instance of your RAC, like locks, logswitches, undo retention and so on. Also if you use ASM you can’t check you ASM1 instance from RAC node 2, because they should be locally accessable only.
I guess you could monitor a RAC via a 3rd server but you need to monitor the instances separately.
We monitor each instance of the database separatly and use the cluster function to map all cluster wide services to the cluster host (you can maybe also use the scl-address of your RAC as name for the host)
This should be all services which are cluster relevant and not node specific (in our case):
Thank you very much. I would follow the Clustered service route now.
This topic was automatically closed 365 days after the last reply. New replies are no longer allowed. Contact @fayepal if you think this should be re-opened.
