RE: OEM and HPServiceGuard
- From: "Walker, Jed S" <Jed_Walker@xxxxxxxxxxxxxxxxx>
- To: Fuad Arshad <fuadar@xxxxxxxxx>, <oracle-l@xxxxxxxxxxxxx>
- Date: Sat, 17 Jul 2010 07:41:16 -0600
Thank you, I’ll take a look at that document and see if it’ll help. From: Fuad Arshad [mailto:fuadar@xxxxxxxxx] Sent: Saturday, July 17, 2010 7:27 AM To: Walker, Jed S Cc: <oracle-l@xxxxxxxxxxxxx> Subject: Re: OEM and HPServiceGuard I'm working on this issue with hacmp which is similar to hacmp . You have 2 options here 1 is to install one agent that monitors the virtual node and is installed in a shared fs that fails over or 2. Use the emcli relocate command as part of the failover to failover the resources from one grid control aget to another . Metalink doc 577443.1 explains the procedures and it is available in 10.2.04+ grid control Fuad On Jul 17, 2010, at 7:13 AM, "Walker, Jed S" <Jed_Walker@xxxxxxxxxxxxxxxxx> wrote: Does anyone have experience with using OEM and HPServiceGuard? We have OEMGC in place, but we have many systems that use HPServiceGuard for failover. The problem is that OEM deals with the physical server address and not the virtual address. This morning we had a server fail. ServiceGuard automatically failed over to another server and the listener, databases, and OEM agent all came up quickly on the new server and we were back in action without any need to touch things. That is, except for OEM, which kept spitting out alerts that the listener and db were down even though they weren’t (any more, just on another server in the cluster). I thought I could reconfigure them to point to the new server (or preferably the virtual host name) but couldn’t. I finally had to put the original server in blackout and now will have to figure out why the agent on the new server isn’t reporting in to the repository. Thanks for any help, Jed.