Resource Manager

This section describes specific issues with the Resource Manager (RM) that might require troubleshooting.

Failover Does Not Work

Problem

The Resource Manager is configured for failover but it is not working.

Resolution

From the command line, issue the $InstallationRoot$/bin/NLB.bat<enable|disable> <cluster node ID> command to see if the traffic can be redirected manually.

• If traffic can be redirected manually, it is an RM issue. Check the configuration options in the RM cluster section to ensure that the IP address and port numbers that are specified for each cluster member (1 and 2) are reachable from the other RM host (1 is reachable by 2 and vice-versa).

  Specifically check the following configurations:

  1. Ensure that the TCP port that is configured in the cluster section for cluster members 1 and 2 is open and is in a listening state. Verify this by running the netstat command.
  2. Ensure that the you can ping the IP addresses that are specified in the cluster section for cluster members 1and 2 from each of the RM hosts.
  3. If the RMs are installed on Windows, ensure that the IP addresses that are specified in the cluster section for cluster members 1 and 2 do not belong to the NLB-dedicated NIC (where the virtual-IP is defined).
  4. Ensure that the firewall, if enabled, is not blocking the communication between the RMs in the cluster.
  5. Ensure that all cables are properly connected.
• If traffic cannot be redirected manually, the issue is outside of the RM and you must check the entire HA configuration.

Both RMs Are Active, When in Active/Standby Mode

Problem

The Resource Manager (RM) is deployed in active/standby High Availability (HA) mode and both RMs are active.

Resolution

This is an indication that the RM nodes cannot communicate properly with each other. See Steps a to e in “Failover Does Not Work”.

Neither RM is Active, When in Active/Standby Mode

Problem

The Resource Manager (RM) is deployed in active/standby High Availability (HA) mode and neither of the RMs are active.

Resolution

This is an indication that the RM nodes cannot communicate properly with each other. See Steps a to e in “Failover Does Not Work”.