What are the SnapSync limitations in an HA cluster in QuTS hero h6.0.0 and later?
Applicable Products
- QuTS hero h6.0.0 or later
- Snapshot Manager
Overview
Starting from version h6.0.0, QuTS hero supports SnapSync jobs in a high-availability (HA) environment. Specifically, you can create SnapSync jobs where the source NAS is in an HA cluster.
Limitations
- Only scheduled and manual SnapSync jobs are supported. Real-time SnapSync jobs are not supported.
- Only the source NAS can be in an HA cluster. The destination NAS cannot be in an HA cluster.
- You must select a cluster interface as the source network interface for the SnapSync job.
If you select a non-cluster interface, the SnapSync job may fail to reconnect after an HA failover or switchover.TipA cluster interface is a network interface configured with a cluster IP address, which allows external devices to connect to the cluster through the same address no matter which device is the active node at any given time.
For details, see What is a cluster IP address and how do I use it?
What Happens After Switching a Node to Standalone Mode
When you press and hold the physical reset button on the active node, you reset the NAS and at the same time remove it from the HA cluster, turning it into a standalone device. The passive node remains in the HA cluster and automatically takes over as the new active node.
In this scenario, the SnapSync jobs are retained on both NAS devices, and the same job on both devices points to the same destination.
When you edit the SnapSync job on both devices, the destination NAS keeps only the most recent change, no matter which device it came from. This may lead to data inconsistency and synchronization conflicts.
We highly recommend users remove the SnapSync job from one of the NAS devices to avoid potential issues.
