Troubleshoot redundant AppEngines

Most troubleshooting happens in the Operations Control Management Console. For more information about using the Operations Control Management Console, see the Operations Control Management Console User’s Guide.

Certain requirements are validated by the system infrastructure. For example, the order in which you configure an object pair for redundancy is validated.

The following problems can occur when you are using redundancy:

• You can configure an AppEngine for redundancy before configuring its associated WinPlatform. If you do this, you see an error message that the Platform (specifically, the RMC) is not configured yet.

• If the RMC IP Address parameter is not configured in both hosting WinPlatforms, then the configuration state of both Primary and Backup AppEngines changes to Error. You also see a message indicating that the host WinPlatform is not configured with the network adapter required for redundant communications. When the RMC IP Address is configured and the WinPlatforms are checked in, the hosted AppEngines are automatically revalidated and the Error state is resolved. If hosted AppEngines are checked out, they are not revalidated.

• If both Primary and Backup AppEngines are assigned to the same WinPlatform and you try to deploy both engines, both the Primary and Backup fail to deploy. You see a message that the Primary and Backup objects must be hosted by different WinPlatforms. Reassign the Backup object to another WinPlatform and deploy it separately.

• If both the Network Address and RMC IP Address parameters in the WinPlatform’s editor refer to the same network card, you get a warning message when you save the configuration. These parameters must refer to different network cards.

• Before restarting a computer that hosts one of a redundant pair of AppEngines (either the Active or Backup), ensure that the Primary Network is connected. A restart while the Primary Network is disconnected makes the Primary Network bind to the RMC’s IP address. An incorrect redundancy state occurs, indicating that redundancy functionality is good. Any time you restart a redundancy-enabled computer, check for proper network connections afterwards. For more information, see Multiple network interface cards.

• When scripting is used to activate attributes, attributes displayed in the watch window may appear to remain in an initializing state after deployment or engine failover. This happens because the script is capturing the attributes while they are still initializing on one of the redundant engines, before the engines are able to synchronize. The attributes will move to active state, but unless there is retry logic in the script, that information will not be returned. This situation can be avoided by ensuring that your scripts contain retry logic to capture the updated state of the attributes.