OVC IP Failover

[J-Philippe]

Hi to the community,

do you know (I never had the opportunity to test it) if a SimpliVity Stretched cluster can manage 2 OVC IP failover at the same time ?

Shuting down one OVC (only the OVC, the nodes remain online) in each Availability Zone of a stretched cluster.

Didn't find documentation about it.

If someone have some feedback, you are more then welcome.

BR,

Jean-Philippe

[s-]

Hi @J-Philippe 

Whenever you stop SVTFS or shutdown OVC, IP will be claimed by its peer OVC in the cluster irrespective of cluster configuration.

Zoning configuration put the P and S copy of vm on either side in the cluster and considering Arbiter is configured at 3rd site.. You may perform the test on your test setup however make sure you are shutting down only one copy of VM. If you are shutting two OVC from either side and both OVC are owning P and S copy of same VM in that situation VM will go down.

Here is HPE SimpliVity Stretched Cluster Technical Overview.

https://kb.vmware.com/s/article/51462

Hope it helps.

Regards,

Jaipal

[J-Philippe]

Hi Jaipal, thank you.

Already know that KB I was misunderstood a point, that's fine now.

So I can deduce, that loosing 2 OVC will impact VM having P & S on but not the other VM on the remaining nodes, so the "DU" will be VM centric and not for the entire storage, correct ?

[Sanika · Accepted solution]

Hi @J-Philippe 

Nodes in a stretched cluster are distributed across geographically separated locations. This redundancy ensures continued VM operation even if one zone experiences an outage.
For example, if a stretched cluster has Zone A and Zone B, VMs have copies in each availability zone for high availability. If Zone A fails, VMs automatically restart in Zone B. In the event of a simultaneous failure in both zones, data might become unavailable for some VMs depending on the specific circumstances of the outage.

In my opinion, losing 2 OVCs may affect VMs having P & S on, these VMs might experience DU because they rely on the failed OVC for configuration & communication. VMs on the remaining nodes with a working OVC are less affected. Their data resides in separate, healthy storage.
And according to my understanding, DU is more VM-centric, impacting specific VMs that relied on the failed OVCs. The overall storage should remain accessible.

Here are some resources that you may find helpful:
https://support.hpe.com/hpesc/public/videoDisplay?videoId=vtc00000092en_us
https://support.hpe.com/hpesc/public/docDisplay?docLocale=en_US&docId=a00131388en_us

Hope this helps.

Regards,
Sanika.

If you feel this was helpful, please click the KUDOS thumb below. Also consider marking this as an "Accepted Solution", if the post has helped to solve your issue.

[J-Philippe]

Hello Sanika, thank you. 

Perfectly aware about the availability zones with SimpliVity. My concern was more about the extent of the "DU".

I feel more confident now. thank you.

BR,
Jean-Philippe

[Sanika]

Hi @J-Philippe 

That's good to know! Happy to help in any way I can.

If you feel this was helpful, please click the KUDOS thumb below.
Also please consider marking this as "Accepted Solution" if the post has helped to solve your issue.

Regards,
Sanika.