DB and ASCS on same cluster

[Former Member]

Dear Experts

I intend to run both DB, ASCS, ERS and CI on same cluster. Is there any issue while running all critical serveries on same cluster.

DB - Should run as fail over.

[Former Member · Accepted solution]

Hi Priyan

Genreally speaking, you should protect with failover cluster only single points of failures, e.g. DB and ASCS instance, and  install at least 2 applications servers (e.g. Central Instance and Dialog Instance) on two different physical host, as they are redundant components.

And you can install all this on one cluster, at least with Microsoft Cluster.

which clustering solution do you want to use?

if you would use Microsoft Cluster, you get all this (with ERS instance) out-of-the box, supported by SAP, intergated in installation etc.

You can check following white Paper: “MSCS Configuration and Support Information for SAP NetWeaver ’04 and SAP NetWeaver 7.0 Systems”, which you can download from:

http://www.sdn.sap.com/irj/scn/go/portal/prtroot/docs/library/uuid/e0b8cd93-f1e9-2910-c186-86bfff3da...

I hope that helps

best regards

Goran

[Former Member · Accepted solution]

Dear Goran

I supposed to install and use SUSE linux HA cluster.

I want to know is there any issue while running both DB and ACSC|CI on same cluster?

thanks

[former_member189725 · Accepted solution]

There should not be any issue with this configuration . You will have to keep DB+ASCS+CI in 1 cluster resource group.

There should not be any issue .

Regards

Ratnajit

[Former Member · Accepted solution]

Hi Priyan

On SUSE Linux should work. Please check SAP Note  1552925 (Linux: High Availability Cluster Solutions) on this.

On Novel web site http://www.novell.com/sap you can find whitepaper "SAP made High-Available on SUSE Linux Enterprise Server for SAP applications"

Although you can have all cluster resources in one cluster group. I would strongly reccomend NOT to do like this - it will cause more downtime - as problem in ASCS can couse failover of the whole cluster group and all the resources inside - DB in this case (although DB has no problemes), and this will cause more dontime.

Also you do not need to cluster CI - CI is not SPOF (Single Point Of Failure) - just insall one CI and one DI on two different server , and you will be fine

best regards

Goran

[Former Member · Accepted solution]

I fully agree with Goran. Place in resource groups only those resources which really need to be protected against failures. Create different resource groups for ASCS and DB

Regards

Roman

[former_member189725]

From the diagram , it seems you want to club the DB,CI and ASCS in 1 resource group and running on physical server 1 and the entire resource group (DB+CI+ASCS) would failover to the other node . The other node physical server 2  would have ERS running. Also only failing over the DB will not be sufficient , ASCS should also failover as both are single points of failure in an SAP system .

So I suggest you have DB+CI+ASCS in 1 cluster group and fail them over together .

The configuration should not have any issues if the above conditions I mentioned is satisfied.

I suggest you go through this link

http://scn.sap.com/docs/DOC-25454#section36

Regards

Ratnajit