Nimble HF20 Updates

[somoa]

Hello

I just got 2 Nimble HF20s and are alerting for a software update . The arrays are joined together in a group with some volumes being replicated using synchronous. There is also a witness with automatic fail over.

My current understanding is I can click the update button and it will do the rest without any downtime. Is there anything I should do or need to do with them before running the update?

I would be going from 6.0.0.500-1005932-opt to 6.1.1.200-1020304-opt

[Vipin_S · Accepted solution]

If a NimbleOS update is available for your array, you can download it now and install it later. The GUI has an automated procedure to install updates.
The NimbleOS update process takes about 70 minutes per array. During that time, a controller failover and a browser reload occur automatically. The array itself remains online and available throughout the update.

If you have multiple arrays in a storage group, all group arrays are updated, one at a time, to the same version of NimbleOS.

 If your connection to the array drops during the update, you might not be able to re-establish connection until the update is done.

When the update is finished, the new NimbleOS version is listed as the Current version and the previous version is listed as Previous.

The update procedure works on one controller at a time and results in a controller failover. To avoid any data service disruption, make sure that the initiators connected to the array have proper MPIO timeouts configured before performing the software update.

If the HPE Nimble Storage Windows Toolkit (NWT) is not installed on the Windows hosts, be sure to configure timeout values appropriately. See Timeout Values.

[Vipin_S]

6.1.1.200 is the latest available and recommended OS version. You can proceed with that. 
In case of any issues during upgrade please reach out to Nimble Support Team.

HPE Nimble Storage

Support 1-877-364-6253

[J_Andersson]

Hi,

We are in the same situation. Previously i beleive there was a document in Infosight describing the supported way for replicated setups.

I have prevoiusly seen info that The witness should be taken offline before upgrade and that Replication should be paused during i update 

Is there a document describing this? Or could we trust Infosight an proceed with the "update" button once a version of Nimble OS are green and all compatabillity matrix options are checked like Witness version, OS Version and SCM Version?

[somoa]

Hi

I did not pause any of these. I also had support on phone and asked them this. They told me there is no need.

My update failed/stopped half way through but never had any downtime during it. Once issue was fixed it restart and all went fine.

Both witness and replication ran the whole time and did not have any issues with either.

[Mahesh202]

Hi Somoa

It's good to hear that your witness and replication continued to run without any issues during the update failure. However, it's important to note that pausing replication during updates is generally recommended to prevent any potential data loss or corruption. It's possible that the update failure didn't have any impact on your replication, but it's always best to err on the side of caution and follow best practices to ensure the safety and integrity of your data.

In any case, if you have any concerns about the stability or consistency of your replication, it's always a good idea to double-check with the HPE Nimble Tech Support team and consider performing any necessary maintenance or troubleshooting steps to ensure that your data is protected.

Hope this helps.!!

Regards
Mahesh

[somoa]

Thanks. I will pause replication in future updates. 

Does that mean witness can be left running?

[Mahesh202]

Hi Somoa

Yes, the witness can be left running during an update.
rather than pausing the replication, the easiest step would be disabling the ASO (Automatic Switch Over), Proceed with the upgrade and later you may enable it.

hope this helps.!!

Regards
Mahesh.

[Mahesh202]

Hi Johan

"Yes" you can trust Infosight and proceed with the "update"
When an update is performed, a controller failover will be part of the process. This alone would not trigger ASO. To be on the safer side, you can disable ASO (Automatic Switch Over) while performing the update and re-enable after.

you could also refer to the below link: https://infosight.hpe.com/InfoSight/media/cms/active/public/HPE_Nimble_Storage_Peer_Persistence_Deployment_Considerations.pdf

hope this helps.!!

Regards
Mahesh.

[J_Andersson]

Hi,

We contacted Nimble support that told us to proceed with the upgrade and keep ASO och Witness enabled and online. Upgrade did complete successfully.

Thank you! I think there should be clear information about the recomended upgrade procedure in the documentation available on Infosight or even a disclaimer link beside the "update button" to minimize risk of failures.

[Yusuf_KAYA]

Hello, unfortunately I could not actively find a document for the Update matrix. When I found the link, it doesn't work. I wonder how we can control the update process? HF20 available and in latest version
It appears as 6.1.1.300-1028597-opt. How will we know if this version supports the HF20 model?
Additionally, the current version is 5.2.1.800-930936-opt. Is there a direct transition to version 6.1.1.300? Can you post an active guide link without explaining these?

[somoa]

In the release note it has a verified update path for each version. From looking at that table I would say you can.

HPE Alletra 6000, Alletra 5000, HPE Nimble Storage Array OS 6.1.1.300 Release Notes

As for the hardwre models that support a release I am not sure where to find that.  Closest I found was a Validated Configuration Matrix | HPE InfoSight