OE 10.2A01 and OEM 10.2A.
Our typical deployment consists of two Linux servers for RDBMS and AppServer with OE Replication between the primary and secondary/backup server (and two Windows server for the GUI client).
We want to use OEM to monitor the server, databases, AppServers, replication etc.
1.Install OEM and the trend databases on the primary server and use remote monitoring to monitor the secondary server.Con: When there is a failover to the secondary server OEM is no longer active.
2.Install OEM and the trend database on the secondary server and use remote monitoring to monitor the primary server.Pro: When there is a failover OEM is still active.Pro: No performance OEM load on the primary server.Con: When the secondary server goes offline OEM is no longer active.
3.Install OEM on the primary and secondary server.Use OE Replication to replicate the trend database from primary to secondary server (just like the application databases).Pro: Primary server and secondary server are functionally equivalent.
4.Install OEM on a 3rd server.Con: We need addtional hardware (possibly the terminal server although we want the terminal server to be as dumb as possible)Pro: Separation of application and monitoring.
In theory I like option 3 but it looks like this will give us some OEM configuration challenges...Is anyone using option 3 in production?What configuration do you use?