Failover Testing and Recovery in HPE Alletra 9080 with Active Peer Persistence

[Elfegume]

Hello,

I am currently performing a PoC with HPE Alletra 9080 and Active Peer Persistence, and I need some guidance on the following points:

1. How can I manually trigger a failover to test my DR environment with Active Peer Persistence enabled? What is the recommended procedure?

2. If the primary Alletra system is manually powered off, once it is turned back on, will the volumes automatically revert to their original state on the primary site? Or is manual intervention required? If so, what are the necessary steps?

My main concern is understanding which actions I need to perform manually and what Active Peer Persistence is capable of handling automatically. The goal is to validate the environment through Data Services Cloud Console and restore everything to its initial state after testing.

I would greatly appreciate any insights or best practices for this process.

Thank you!

[james783]

Hello,

Absolutely! Let's break down the process of manually triggering a failover with HPE Alletra 9080 Active Peer Persistence (APP) and understanding the automatic and manual steps involved in a failover and failback.

1. Manually Triggering a Failover for DR Testing

The recommended procedure for a controlled failover in an Active Peer Persistence environment involves using the Data Services Cloud Console (DSCC) or the Alletra CLI. Here's a step-by-step guide:

Verify Application Health: Before initiating the failover, ensure that all applications using the replicated volumes are in a consistent state. This minimizes data loss and ensures a smooth transition.
Identify the Primary Site: Confirm which Alletra 9080 system is currently the primary (active) site. You can do this through DSCC or the CLI.
Initiate the Failover:
Using DSCC:
Navigate to the "Volumes" or "Protection" section related to your APP configuration.
Locate the volume or volume collection you want to fail over.
Look for the "Failover" or "Switchover" option. (The exact terminology might vary slightly depending on the DSCC version.)
Select the secondary Alletra 9080 system as the target site.
Confirm the failover operation.
Using CLI (Alletra OS):
Use the appropriate CLI commands to switch the volume's ownership to the secondary site. Consult the Alletra OS documentation for the precise commands. The general idea is to change the volume's preferred location.
Example (This is a conceptual example, please verifiy correct syntax in your Alletra OS documentation):
controlport volume modify -name <volume_name> -preferred_location <secondary_alletra_serial_number>
Verify Failover Success:
Check DSCC or the CLI to confirm that the volumes are now active on the secondary site.
Test application functionality on the secondary site to ensure that the failover was successful.
Host Configuration: Depending on your host multipathing software and configuration, you may need to rescan disks or perform other host-side operations to ensure the hosts recognize the volumes on the secondary site.
2. Automatic and Manual Steps During and After a Primary Site Power Off

Automatic Failover (In Planned or Unplanned Outages):
Active Peer Persistence is designed to automatically fail over volumes to the secondary site in the event of a primary site failure, including a power outage.
The Alletra systems continuously communicate and monitor each other. If the primary system becomes unavailable, the secondary system automatically assumes ownership of the volumes.
The hosts, with proper multipathing configuration, should automatically switch to the active paths on the secondary site.
Failback (Primary Site Recovery):
Manual Intervention is Typically Required: When the primary Alletra system is powered back on, the volumes will not automatically revert to their original state. You need to initiate a failback operation.
Failback Procedure:
Verify System Health: Ensure that the primary Alletra system is fully operational and healthy.
Initiate Failback using DSCC or CLI:
Similar to the failover process, use DSCC or the CLI to initiate a "Failback" or "Switchback" operation.
Select the primary Alletra system as the target site.
Example CLI concept:
controlport volume modify -name <volume_name> -preferred_location <primary_alletra_serial_number>
Verify Failback Success:
Confirm that the volumes are now active on the primary site.
Test application functionality to ensure everything is working correctly.

Best Regards

[Elfegume]

Ok, thanks for the clarification, but I still have some doubts.

1. Let’s simulate a failure of the primary system, but we can’t power it off, as there are other operational volumes that cannot be affected. We just want to simulate it on a specific volume. Doing a switchover doesn’t fully convince me as a test. Shouldn’t we do a manual failover on that test volume?

2. Then, with Active Peer Persistence active on that test volume, shouldn’t I perform a recover and then a reverse to restore everything to its original state?

3. Is failback essentially the same as executing recover and then reverse?

Thanks.

[Mr_Techie]

- Manual Failover (Planned)

• Use DSCC to failover just the test volume or volume group to the DR site.
• This promotes the DR copy to read/write and demotes the primary to standby.
• Simulates a localized failure.   

- DR Validation

• Confirm hosts at the DR site can access and perform I/O on the test volume.
• Monitor performance and replication health during the test.

- Recover

• Once testing is complete, perform Recover to re-establish replication from DR → primary. and changes made during DR test are fully synced.

- Reverse

• Then perform Reverse to make the original site active again
• Restores the pre-test replication direction and volume access.

Fyi, If DSCC offers a Failback button, this typically performs both Recover + Reverse automatically.

I hope this give some insights, let me know if this help to resolve 

[support_s]

Hello,

Let us know if you were able to resolve the issue.

If you are satisfied with the answers then kindly click the "Accept As Solution" button for the most helpful response so that it is beneficial to all community members.

Please click on "Thumbs Up/Kudo" icon to give a "Kudo".