Skip to main content
Migration Notice
We're migrating documentation from the old portal into this one. Some things may look a little different or out of place in the meantime — we know, and we're working to get it right. If something's unclear or doesn't look right, let us know.
Version: 4.4.0

Qumulo Failover Overview

Introduction​

Failover in a Qumulo storage environment ensures data availability and continuity in the event of an infrastructure failure. Unlike VAST, which supports failover at three levels of granularity (Tenant, VIP Pool, and Protected Path), Qumulo failover operates at a single, consistent unit: the Replication Relationship.

A Replication Relationship pairs a source path on one cluster with a target path on another. Eyeglass discovers these relationships automatically and uses them as the atomic unit for both configuration replication and failover.

Configuration Replication​

For each Replication Relationship Eyeglass discovers, it creates jobs that keep configuration data synchronized between the source and target clusters. For Qumulo, configuration replication covers:

  • SMB shares
  • NFS exports
  • Snapshot policies

These jobs run automatically every 5 minutes from the Running Jobs section of the Jobs window. You can also trigger a configuration replication job manually — see Common Operations for the general job-management workflow, which applies the same way across storage platforms.

Replication Relationship Failover​

Overview​

  • Failover happens per Replication Relationship — you select the specific relationship(s) to fail over, rather than a broader construct like a tenant or pool.
  • Both controlled and uncontrolled failover are supported. See Execute Failover with DR Assistant for what each Failover Option does and which options apply to Qumulo.

Post-Failover Considerations​

  • Remount required: SMB shares and NFS exports must be remounted by clients after a failover completes — this is not automated.
  • SPN considerations: If your environment uses DNS A records or CNAMEs to direct clients to the active cluster, any associated Service Principal Names (SPNs) in Active Directory must be moved to the target cluster manually as part of the failover.

See Also​