Search code examples
marklogicmarklogic-10

One of the file system (mounted drive) failed in Marklogic server


One of the file system (mounted drive) failed in Marklogic server. We configured Marklogic cluster with high availability (with replica forest) One of the file system (mounted drive) failed in one host but now replica forest is running as master and application is running without any issues. how can I mount my new drive if I mount new drive it will sync with replica forest.?


Solution

  • MarkLogic will automatically failover and open HA replicas, but failing back is a manual decision. This is to avoid thrashing and constant failover events if the host or infrastructure is not stable. Though, it might be nice to have the option to configure thresholds for automatically flipping back to primaries if it does appear to be stable.

    If you have resolved the issue(s) on the host that had the drive failure and are ready to bring it back online, you can (re)mount that drive. When those mounts and forests become available, they will attempt to sync with the HA replicas that are currently serving traffic and acting as primary.

    You should see messages about forests synchronizing (if you have local disk failover enabled, with replicas) and changing states from wait or async replication to sync replication. Once all the forests are either 'open' or 'sync replicating', then your cluster is fully operational with the correct number of hosts.

    When the primary forests are all sync replicating, then you can restart the HA replica forests to fail-back and have the primary forests take over.

    You could use the script provided in this knowledgebase article to flip the open replicas: https://help.marklogic.com/Knowledgebase/Article/View/scripting-failover-flipping-replica-forests-back-to-their-masters-using-xquery