Help us improve your experience.

Let us know what you think.

Do you have time for a two-minute survey?

Navigation
Guide That Contains This Content
[+] Expand All
[-] Collapse All

    Troubleshooting Server Replication

    This section describes how to troubleshoot server replication.

    Loss Of Connectivity to the Cluster

    When the PSM server loses connectivity with another member in the cluster, the server raises a 'PSM cluster node connection lost' alarm, identifying the member with which connectivity has been lost. Typically, the servers at both ends of the connection will raise the alarm, signifying a loss of connectivity with their counterpart.

    When this occurs, it is best that you do not change any replicated data on the local server or on any server with which connectivity has been lost. If you do make changes, you may have to manually reconcile these changes later. This becomes particularly complicated if there are more than two servers in the cluster.

    Loss Of Synchronization with a Cluster Member

    When the PSM server loses synchronization with another member in the cluster, the server raises a 'Synchronization failed against cluster node' alarm, identifying the member with which synchronization has been lost. Synchronization is declared lost when the remote member does not respond to a synchronization request. A server sends out synchronization requests on restart and on manual command.

    When this problem occurs, perform a manual synchronization (Synchronizing replicated data manually).

    Re-establishment Of Connectivity to the Cluster

    When the PSM Server re-establishes connectivity with the other member(s), the server clears the 'PSM cluster node connection lost' alarm(s).

    If you did not make any changes to the replicated data on the local server or on any of the servers for which connectivity was lost, then you do not need to take any action. All servers have the correct replicated data and are synchronized.

    If you did make changes to the replicated data on the local server and/or on any of the other servers for which connectivity was lost, then you will need to take the following action:

    Synchronizing Replicated Data Manually

    Use this procedure to manually synchronize replicated data with other servers in a cluster.

    Before executing this command, ensure the following:

    • When running with two servers in the cluster, ensure the other server has the correct replicated data.
    • When running with more than two servers in the cluster, ensure all other servers have the correct replicated data. If not all of the other servers have the correct replicated data, then do not run this command.

    Perform manual synchronization of replicated data on the local server when you want the local server to retrieve and adopt replicated data from the other servers. This is typically performed after recovering from a loss of connectivity, but can be done at any time.

    Servers in a cluster do not automatically resynchronize with each other after recovering from a loss of connectivity because the servers have no way of knowing which server has the correct replicated data. Instead, you have to decide which server has the correct replicated data, and then manually resynchronize all the other servers with that server.

    1. To force the local PSM server to manually synchronize with all servers in the cluster, select Tools >Synch Replication.

      Note: This option is not available for selection if server replication is disabled. See the proNX Service Manager Installation and Administration Guide for information on how to enable server replication.

      The local PSM server retrieves replicated data from all other servers in the cluster, and overwrites its own replicated data with the retrieved data.

    Restarting the Cluster

    Use this procedure to restart the cluster with the correct replicated data.

    In order to execute this procedure, you must have access to the Linux operating system shell on the PSM Server.

    1. Decide which server in the cluster has the correct replicated data. Leave this server running.
    2. Stop the PSM server application on all the other servers.
    3. Start the PSM server application on each server one by one.

      Note: Ensure each server is up and has the correct replicated data before bringing up the next server.

    4. Proceed until all the servers are up and have the correct replicated data.

      Note: For information on how to stop and start the PSM Server, see the proNX Service Manager Installation and Administration Guide.

    Modified: 2017-03-29