Point-in-Time-Recovery (PITR) Fails

This article applies to JSDR software version 4.2.x.

Problem

The PITR recovery process fails and the task reports the error, “Task is already running.“

Solution

1. Validate if the protected VMs reside on a shared datastore or local datastore.
    
2. Protected VMs on a local datastore may fail PITR recovery under certain conditions.
   • If the protected VMs reside on a local datastore and the datastore is accessible from any other host in the cluster, there is a possibility the RVMs and RocVA may be created on a host different from the original VMs. In such case, the host would not be able to access the original local datastore and PITR recovery can fail.
      
3. If all protected VMs are located on shared datastores, then selecting any datastore for PITR recovery should work. This is feasible because RVMs created on any host will maintain their ability to access the original disks situated in shared storage.

For the Point-in-Time Recovery (PITR) feature to function reliably, it is necessary to have all JetStream DR protected VMs located in shared datastores.