Initial Hyperion EPM Troubleshooting

Troubleshooting Hyperion can be a difficult task depending on what’s happening in an environment and/or what components are installed.  There could be complex environments where almost every Hyperion product is set up across multiple servers, or it could be a simple Planning environment with 2-3 servers.  This is not meant to be a comprehensive guide, but to give a general idea of things to look for when investigating errors or performance issues.

When troubleshooting, if you’re not intimately familiar with all the random error codes that can pop up, the best place to start is with the service logs which can be your best friend.  These are located at Middleware Home\user_projects\epmsystem#\diagnostics\logs\services.

Middleware Home is where the installation was specified, typically something like D:\Oracle\Middleware, or C:\Oracle\Middleware.  Some services don’t have service logs like HFM server, so you’d go to Middleware Home\user_projects\epmsystem#\diagnostics\logs.

If the service logs aren’t showing any errors or not helpful, look at the WebLogic logs located at Oracle Home\user_projects\domains\EPMSystem\servers.  Under the server folder, go into each subfolder and then the logs folder (if it exists).  You may see the same errors or find errors that weren’t in the service logs.

After going through the logs, and if the error doesn’t help you determine what the root cause of the problem is, keep in mind that Hyperion is composed of multiple parts and layers.  In a complex environment containing multiple servers, the problem could be with the physical or virtual server, a 3rd party component such as Microsoft’s IIS, the network or database.

A common scenario is that you go through the logs and see errors that have JDBC in them.  This can mean, for whatever reason, Hyperion lost connection to the database.  You would want to investigate if the database was restarted as part of maintenance if there was a network failure, or something happened on the server Hyperion is running on.

Another simple thing to check is to see if Hyperion services are running.  If one or more crashed on a server, then check to make sure the server has enough available memory and/or disk space.  Both, available memory and disk space, can cause services to crash since they are fighting for available resources.  Running out of disk space can also have other negative side-effects such as corrupting the SSL Oracle Wallet or Essbase server configuration files.

On a Windows server, If Hyperion isn’t performing well, going through the Oracle tuning guides and documentation and ensuring that values are set correctly get easily increase performance and stability.  Usually, these values are set and changed in the Windows registry and IIS.

The information above is to assist with the first round of troubleshooting in a Hyperion environment.  This is by no means all-inclusive and there are many other root causes and solutions that may be the answer.  However, the list above is a great place to start and will capture the most common and highest number of issues that can cause a system down in a Hyperion environment.