ECC Upgrade 4.7 tlo ECC6.0 in HACMP environment

[Former Member]

Hi All,

I am currently doing upgrade for one of my client R/3 4.7 to ECC6.0. The R/3 is running on AIX 5.3 with Oracle 10g (10.0.2.0.4).

There production server is running on Highavailability (HACMP) cluster.

Kindly let me know if any body has done the upgrade on HACMP environment and what all precaution or additonal steps we have to follow.

Thanks in advance.

[former_member126060 · Accepted solution]

Hello Rajesh,

Here are the important points need to be taken care for upgrade in HIgh availibility environment. There are very well documented in the latest upgrade guide. Here is the compiled list :

The upgrade of high availability systems can only be performed with upgrade strategy downtime-minimized.

The upgrade program can handle distributed configurations where your ASCS instance is installed on a separate host (for example, in a high availability environment). As a prerequisite, the SAP kernel directory must have the new layout (/usr/sap/<SAPSID>/SYS/exe/uc/<platform>) and the program sapcpe needs to be configured. For information about upgrading a high availability system, contact your HA solution implementation partner.

The upgrade process for a high availability system is as follows:

1. You prepare the disabling of the cluster failover capabilities and, if necessary, change your configuration and clean up profiles (see Preparing the Upgrade of Distributed and High Availability Systems.

You will be prompted you to disable the cluster failover capabilities to allow the upgrade program to stop and restart clustered instances.

2. You start the upgrade program on the central instance host.

3. The upgrade program checks the profile directory and, if the ASCS instance is installed on a remote host, assumes that the upgrade is running in a distributed or high availability environment.

4. At the beginning of the downtime, the upgrade program prompts you to stop the remote ASCS and Enqueue Replication Server (ERS) instances manually.

(Phase REQ_ASCS_STOP).

5. During the downtime, the upgrade program temporarily reconfigures the central instance to include the ASCS instance.

6. At the end of roadmap step Downtime, the upgrade program prompts you to install the latest SAP kernel, start the ERS of the ASCS instance and restart the ASCS instance.

(Phase REQ_ASCS_START)

If you run the ASCS instance and the ERS instance on separate hosts (for example, in a high availability environment), you are asked to do the following:

- On the host with the ASCS instance, install the latest SAP kernel.

- Restart the ERS of the ASCS instance and then restart the ASCS instance.

If your system is installed in a high availability switch-over environment, make sure that the failover capabilities of the cluster switch-over software are disabled during the downtime. This is to prevent unwanted failover actions during stop actions of the upgrade procedure.

Although the toll will capture if you are working with HA environment and if found you will get the prompts to take actions in the course of the upgrade but go through the upgrade guide first and make yourself comfortable.

Best Regards

Niraj

Edited by: Niraj Kumar Soni on Apr 25, 2010 3:14 PM

Edited by: Niraj Kumar Soni on Apr 25, 2010 3:16 PM

[Former Member]

Thanks Niraj for your detail answer. It will be very helpful for me. I have too doubt:

1. Disable the cluster failover.

Do we require to bring the database and CI on the same box since our production is running on cluster.

5. During the downtime, the upgrade program temporarily reconfigures the central instance to include the ASCS instance.

As mention during downtime of upgrade the CI includes the ASCS instance also. Once the upgrade is compelte what changes we have to make to seperate the CI and ASCS instance.

Thanks once again.

[former_member126060]

Hello Rajesh,

Here are the answers :

1. Disable the cluster failover.

Do we require to bring the database and CI on the same box since our production is running on cluster.

==> Yes

5. During the downtime, the upgrade program temporarily reconfigures the central instance to include the ASCS instance.

As mention during downtime of upgrade the CI includes the ASCS instance also. Once the upgrade is compelte what changes we have to make to seperate the CI and ASCS instance.

==> You will not have to do anything. When the tool prompts you to stop your ASCS you should stop it. for upgrade purpose a temporary ASCS is created, which will be later dropped.

I hope that makes clear.

Best Regards

Niraj

Edited by: Niraj Kumar Soni on Apr 27, 2010 12:14 PM

[former_member126060]

Hello Rajesh,

Discard the above message, somehow it was not properly formatted. I tried it 2 times :

Here are the important points need to be taken care for upgrade in HIgh availibility environment. These are very well documented in the latest upgrade guide. Here is the compiled list :

The upgrade of high availability systems can only be performed with upgrade strategy downtime minimized.

The upgrade process for a high availability system is as follows:

1.You prepare the disabling of the cluster failover capabilities and, if necessary, change your configuration and clean up profiles (see Preparing the Upgrade of Distributed and High Availability Systems.

You will be prompted you to disable the cluster failover capabilities to allow the upgrade program to stop and restart clustered instances.

2.You start the upgrade program on the central instance host.

3.The upgrade program checks the profile directory and, if the ASCS instance is installed on a remote host, assumes that the upgrade is running in a distributed or high availability environment.

4.At the beginning of the downtime, the upgrade program prompts you to stop the remote ASCS and Enqueue Replication Server (ERS) instances manually.

(Phase REQ_ASCS_STOP).

5.During the downtime, the upgrade program temporarily reconfigures the central instance to include the ASCS instance.

6.At the end of roadmap step Downtime, the upgrade program prompts you to install the latest SAP kernel, start the ERS of the ASCS instance and restart the ASCS instance.

(Phase REQ_ASCS_START)

If you run the ASCS instance and the ERS instance on separate hosts (for example, in a high availability environment), you are asked to do the following:

- On the host with the ASCS instance, install the latest SAP kernel.

- Restart the ERS of the ASCS instance and then restart the ASCS instance.

If your system is installed in a high availability switch-over environment, make sure that the failover capabilities of the cluster switch-over software are disabled during the downtime. This is to prevent unwanted failover actions during stop actions of the upgrade procedure.

Although the tool will capture if you are working with HA environment and if found you will get the prompts to take actions in the course of the upgrade but go through the upgrade guide first and make yourself comfortable.

Best Regards

Niraj

Edited by: Niraj Kumar Soni on Apr 25, 2010 3:22 PM