I've scoured the Oracle documentation and not been able to find a definitive answer to this question, how many Oracle management agents can you have on a single managed host target? Can you for example run an 11g agent and a 12c agent simultaneously (whilst migrating to 12c without upgrading the present OEM instance) If so can you run two 12c agents, each pointing to a different OMS? I’ve seen examples of an agent for each database instance on a host managed target, so a host with three Oracle database instances has three Oracle Management agents all pointing to the same OMS with each agent is monitoring just one of the database instances. There must be resource considerations here for memory, cpu and network traffic etc? I’ve always understood it to be one Oracle Management Agent per managed host
You may run multiple agents on a single host, but it is not a best practice to do so. The upgrade scenario you describe is actually one reason you might want to do this as a temporary strategy. Each agent deployed on a host must upload to a different OMS. The example you cite of multiple agents, all pointing to the same OMS, is not a supportable configuration.
Many thanks for the reply, I don't suppose you've any evidence that might suggest that running multiple agents pointing to different OMS's can cause performance issues do you? I just wondered if there was a chance they could contend with each other for resources.
I don't think there's any data on impact or performance since it's not a recommended configuration. If you have host/configuration monitoring enabled on both agents, they will be running the same queries/scripts against the host/inventory, etc. so they will likely have contention at some point. If you monitor the same databases, you will definitely see contention and increased overhead as there will be multiple agents running the same queries. You might also see problems w/ any of the server generated alerts (tablespace, etc) if you change thresholds as the agents send that to the database, and then the database will be out of sync with the other agents.
What exactly is the goal for having multiple agents? Is it short term or long term?