Print Email PDF

Using Failover with Replication in Qumulo Core 2.12.0 (or Higher)

IN THIS ARTICLE

This article explains how to use failover with replication in Qumulo Core 2.12.0 (or higher).

REQUIREMENTS

  • Administrative privileges
  • Source and target clusters, both with Qumulo Core 2.12.0 (or higher)

If you run a version of Qumulo Core lower than 2.12.0, see Replication: Failover and Failback with 2.11.4 and Below.

DETAILS

There are two types of failover, planned and unplanned. Replication helps you recover from either scenario. The following sections explain performing planned and unplanned failover on your clusters.

To Perform a Planned Failover to a Secondary Cluster

The following steps explain how you can perform a planned failover with replication.

  1. To ensure that the target directory is in a point-in-time consistent state that matches the last source snapshot, do the following.
    1. Set the source directory on your primary cluster to read-only.
    2. To ensure that no writes are lost, wait for a new replication job to start and finish.
    3. To ensure that a job is finished, verify that any blackout windows are disabled and then check the relationship's recovery point time, ensuring that is after the time that you set the source directory to read-only.
  2. In the Web UI of your secondary cluster, click Cluster > Replication.
  3. On the Replication page, under Actions, click Screen_Shot_2022-02-24_at_6.19.26_PM.png > Make Target Writable...
  4. In the Make Target Writable dialog box, click Yes, Make Target Writable.
  5. Wait for your directory to revert to its last recovery point.
    To monitor progress, click Details.
  6. When the Status column displays the following message, the target directory is consistent and writable.
    Disconnected and target writable. Reverse or reconnect using the relationship actions under the actions menu.
  7. If the following configuration doesn't already exist on your secondary cluster, migrate it:
    • NFS exports
    • SMB shares
    • AD/LDAP servers
    • Snapshot policies
    • Quotas
  8. Remount any clients that require access to your secondary cluster and were previously connected to your primary cluster.

The planned failover process is complete.

To Perform an Unplanned Failover to a Secondary Cluster

If the previous replication was incomplete before the primary cluster became unavailable, it might have left the target directory in an indeterminate state with some files truncated or inconsistent.

The following steps explain how you can perform an unplanned failover with replication by bringing the target directory to a determinate state using the last successful replication recovery point to let you use it as the source directory.

  1. In the Web UI of your secondary cluster, click Cluster > Replication.
  2. On the Replication page, under Actions, click Screen_Shot_2022-02-24_at_6.19.26_PM.png > Make Target Writable...
  3. In the Make Target Writable dialog box, click Yes, Make Target Writable.
  4. Wait for your directory to revert to its last recovery point.
    To monitor progress, click Details.
  5. When the Status column displays the following message, the target directory is consistent and writable.
    Disconnected and target writable. Reverse or reconnect using the relationship actions under the actions menu.
  6. If the following configuration doesn't already exist on your secondary cluster, migrate it:
    • NFS exports
    • SMB shares
    • AD/LDAP servers
    • Snapshot policies
    • Quotas
  7. Remount any clients that require access to your secondary cluster and were previously connected to your primary cluster.

ADDITIONAL RESOURCES

Replication: Failover and Failback with 2.11.4 and below

Replication: Failover and Failback with 2.12.0 and above

Replication: Version Requirements and Upgrade Recommendations

Snapshots: Deep Dive

QQ CLI: Replication

 

Like what you see? Share this article with your network!

Was this article helpful?
0 out of 0 found this helpful

Comments

0 comments

Please sign in to leave a comment.

Have more questions?
Open a Case
Share it, if you like it.