New release Simplivity V5 upgrade process

[kct44]

The new HPE Simplivity V5 release, which involves more changes than previous upgrades.

The change to SVT 5.0 implies a change of Ubuntu version from 18.04 to 22.04 and also a renaming of Simplivity's network interfaces => All scripts that may call upon these ethxx interface names must be adapted. (Supervision, deployment ..).

Apart from the Interropability doc, the simplivity upgrade planner...do you have a detailed upgrade process? Or does the upgrade manager update the omnistack to 22.04 and change the interfaces to svtethx automatically? Is there a script to run at the end of the upgrade?

Thanks for your feedback,

Julian,

[alg204]

Used Upgrade Manager to upgrade to 5.0 and it took care of everything. Didn't have to do anything afterwards. Just in case, I ran all upgrades on one node and waited a bit. No issues, so then I ran the rest.

[kct44]

Thank you for your feedback. I thought it was necessary to run another script after the update.
So I'll be able to start a 1st update on a small 2-node cluster.

See you soon,

Julian,

[s-]

Hi @kct44 

Please note Upgrading the Omnistack Virtual Controller (OVC) to version 5.0.0 could potentially fail at 97 percent. When the failure occurs, the OVC will roll back the upgrade to the previous version and the services (svtfs) restart will fail.

SimpliVity nodes that were deployed on version 3.7.9 or earlier and then upgraded to later versions are exposed to this issue due to an incorrect PostgreSQL configuration parameter.

Here is the customer advisory for Simplivity upgrade 5.0 :

https://support.hpe.com/hpesc/public/docDisplay?docId=emr_na-a00138760en_us

Hope it helps.

Regards,

Jaipal

[kct44]

@s- Hello,

Thank you for your feedback. I hope it's the only advisory for this upgrade process :-).

And how much do you estimate for this upgrade per FW, VMW and SVT patch node?

Have a nice day,

Julian,

[s-]

Hi @kct44 

It should take approx. 1 hour to 1.30 hour per node, considering you are upgrading OVC, ESXI and firmware etc.

Hope it helps.

Regards,

Jaipal

[Convergo]

@s- I noticed infosight is also referring to this article, even when we use a 4.2 cluster. I guess we need to apply this advisory proactively or is it a false positive Infosight warning ?

[kct44]

Hello,

I can confirm that the recommendation must be followed. In addition, if you have an external storeonce bay connected to the federation, you should also follow the note below:

https://support.hpe.com/hpesc/public/docDisplay?docId=emr_na-a00139083en_us&hprpt_id=ALERT_HPE_3066216&jumpid=em_pom8nu6hj_aid-521053889

Julien,