EHP6 update on Windows cluster failover issue

[former_member286941]

Hi,

We are doing EHP6 update from ERP6 EHP5 on Windows cluster 2008 R2, SQL server 2008.

As per SAP recomendations the cluster nodeA needs to be disabled after moving all the cluster resources to NodeB before actual downtime starts.

what is the possibility to proceed in downtime if the cluster resource DB is on NodeA and ASCS, MSDTC on NodeB, and we are not able to move the resources.

If we have no other option except this situation, then when and what should be taken care?

Thanks in Advance.

Regards

Munish Jindal

[Reagan]

Hello

Before the downtime starts, you will need to disable the cluster failover capabilities and move the SAP<SAPSID>HA resources to the central instance host.

This is clearly mentioned in the upgrade guide.

what is the possibility to proceed in downtime if the cluster resource DB is on NodeA and ASCS, MSDTC on NodeB, and we are not able to move the resources

Is there a technical difficulty in doing this ?

Good Luck

RB

[former_member286941]

if this happens, failover fails from one node to other before dowtime, and we have to start  what could be further line of action.

Regards

Munish

[former_member286941]

The reason is, Last time dev enviorment has problem in doing this, this time in production we don't have much time..

thats the reason if anyway its asking for more then expected time to fix cluster issue, its will delay the deadline.

regards

Munsih

[former_member215961]

Hi,

Sap recommentation is made by the well running of SUM upgrade tool.

If you start the SUM at your configuration you must finish the upgrade at the same.

Cheers

[former_member286941]

Hi Sargan,

Thanks for reply.

I don't get your answer. what do you mean by "

If you start the SUM at your configuration you must finish the upgrade at the same."

My question is little different,

"

what is the possibility to proceed in downtime if the cluster resource DB is on NodeA and ASCS, MSDTC on NodeB, and we are not able to move the resources.

"

Regards

[former_member215961]

I mean that if SUM is started when NodeA has ASCS instance and NodeB has DB instance if you move the resources SUM maybe cannot find somenthing...

Anyway your question was that the possibility of move forward because you cannot move all the cluster resources at the same node. Technically I don't know any inconvenience.

Cheers