A few days ago I received a request to restart a set of dependent Hyperion services every Sunday. This had to be done because; “That’s what Oracle support said is the fix to one of our TAR’s”. Although this server has 3.75 GB of RAM and had never used more than 1.2 GB of RAM. Oracle claims that memory issues are causing a service to fail. This failure causes logons to fail. There are no events in the eventlogs or error information in any Hyperion logs that we can find. However, If you restart the services, then the problem goes away. So that is Oracle support’s answer, restart the services; “case closed”.
I do not call this a solution, but a kludge. Our Hyperion activity is relatively light. I guess Oracle’s answer to their very large and active customers must be to restart these services every 15 minutes. Hmmm, no let me think a second, the Hyperion product is financial planning software, with support answers like this, should I be concerned that this software might be calculating the wrong results? I wonder if I need to restart the other services frequently, just to be sure they calculate correctly?
Ok, my sarcasm and diatribe aside, I decided to create a PowerShell script and run it as a scheduled task.
continue reading » »