cancel
Showing results for 
Search instead for 
Did you mean: 

SAP start-stop if cluster is down

Former Member
0 Kudos

Dear Experts,

We have cluster environment for SAP ECC6.0 with 2 nodes i.e. CI&DB. ASCS<Instance_Number> and DVEBMGS<Instance_Number> are in CI node.

Imagine,If cluster goes down/switch off, how can we start the application. Is it possible?

Accepted Solutions (0)

Answers (1)

Answers (1)

Former Member
0 Kudos

Hi Sumit,

  In case if your environment is Windows, kindly have a look at the below link

http://www.sdn.sap.com/irj/scn/go/portal/prtroot/docs/library/uuid/40533d65-572a-2f10-6183-a4dd7f21d...

In case if your Environment is Unix or linux, please have a check on the below link...

in example, we have node A as db server and this server (node A) that is clustered with apps1 server (node B)

also we have another one apps2 server.

assume that cluster, active node and package is active on db server (node A)

stop sap in cluster :

- stop sap in all apps server (in this example apps2 server)

- stop os collector on apps2 server

- stop sap in node B (in this example apps1 server)

- stop os collector on node B

- stop sap in node A (in this example db server where cluster is run on it)

- stop os collector on node A

- login as oracle user

- stop listener

- exit

- stop the cluster

- view cluster status to verify cluster have been stopped succesfully.

start sap in cluster environment

(different cluster design will result different behaviour, in our environment starting the cluster will call the script to automatically start the cluster, start listener, start db server - node A, start apps1 server - node B and start apps2 server)

but let's assume that your setting needs to manually start everthing. so the steps are :

- starting the cluster

- login as oracle user, and start the listener

- start the db server (node A) -- it will automatically start oracle database first and then starting the sap services

- start the apps1 server (node B)

- starts another apps server (apps2 server)

I hope it helps..

Please check and provide your feedback.

Thanks and Regards,

Vimal

Former Member
0 Kudos

Dear Maria,

My query  is simple Can we start SAP application manually if cluster gets down. You are suggesting me start-stop of SAP with cluster on.

SAP landscape is on linux cluster.

Former Member
0 Kudos

Yes Sumit...

It is possible to start the SAP system even if one cluster node is down....

This is one of the features of SAP Cluster

Former Member
0 Kudos

  Hi Maria,

Yes, I know, if cluster fails on 1 node, SAP switch on other node.

I am asking, If cluster services completely goes down, then How can we start the application manually

Former Member
0 Kudos

I am afraid that in that case, it is not possible....

Former Member
0 Kudos

Dear sumit,

    Why dont we try to OS level?

We can execute startsrv.exe and saposcol.exe and share server files at network.

server will up.

Regards,

Prabaharan B

Former Member
0 Kudos

Dear Maria,

http://help.sap.com/saphelp_nwpi711/helpdata/en/47/1d6feeff6e0d46e10000000a15536/frameset.htm

From above information, It means  if only ASCS  is  running on switchover cluster, and DB instance  & CI instance running on individual hosts,  then we can perform manual start-stop (in case of cluster shutdown). 

In our case, ASCS & CI is running on 1 node and DB is on other node.  Need to understand how can we made available only ASCS services on switchover  cluster.

Please suggest....

Former Member
0 Kudos

Hi Sumit,

Please open the attached picture for your reference.

It is more or less similiar to your scenario.

So here is the procedure.

1. Install ASCS in the shared disk.. So that even if Node A fails, shared disk that has the ASCS will move to Node B.

2. Install DB in the shared disk.. Same as ASCS. In the event of Node A failure, Database will be moved to Node B.

3.Install CI in the local disk of Node A.. it shouldnt move to the other node in the event of node failure.

4.Install DI in the local disk of Node B .. It shouldnt move to the other node in the event of node failure.

So what happens,

Suppose Node A fails, ASCS and DB will be moved to Node B so that ASCS will be running in Node B along with DI and vice versa .

I hope the picture helps you to understand and have an idea .

Please check and provide your feedback..

Thanks and Regards,

Vimal