So we’re undergoing a massive push to virtualization at a co-location facility (rented datacenter, if you will), and my first smallish production Oracle database got migrated. As a result, the application server (Windoze) and the 11g Oracle database (OL) got new IP addresses. Always scary, but in the end I knew Oracle would be less problematic than the application server because Oracle doesn’t really have strong ties to IP addresses.
The move went great except for one thing: when I brought up my database on the new subnet, the agent was acting all crazy. It reported everything was up (meaning the listener, host and database) but I couldn’t drill into anything via grid control. It gave me a succinct, useless error:
I’ll type it here so there’s some change Google can index it: Database Error The Network Adapter could not establish the connection
Well this didn’t mean anything, so I figured I’d leverage my support and open a ticket with Oracle. The analyst didn’t exactly nail the issue, but as they were asking for logs I took a peek and noticed this in the $AGENT_HOME/sysman/log/emctl.log file:
28319 :: Fri Oct 11 06:15:12 2013::AgentStatus.pm:Processing status agent
28319 :: Fri Oct 11 06:15:12 2013::AgentStatus.pm:emdctl status returned 3
30263 :: Fri Oct 11 07:15:35 2013::AgentLifeCycle.pm: Processing status agent
30263 :: Fri Oct 11 07:15:35 2013::AgentStatus.pm:Processing status agent
30263 :: Fri Oct 11 07:15:35 2013::AgentStatus.pm:emdctl status returned 3
The fix? emctl secure agent
After I secured it, I logged into EM 11g (yeah, I’m still not on EM 12c, it’s complicated) and got right into the tabs and pieces I wanted to see. My best guess is that the agent became unsecured as a result of the IP address change, and I merely had to re-secure it to make everything work again on the new IP.