cancel
Showing results for 
Search instead for 
Did you mean: 

SAP High Availability installation

Former Member
0 Kudos

Hi Exports,

My Client wants to go with HA with SQL server 2012 by using following structure

i have 4 OS

1 SAP client 1

2 DB instance 1

3 SAP client 2

4 DB instance 2

and he wants if my SAP Client 1 gets crashed then SAP client 2 will get connected automatically, if my DB instance get crashed then SAP Client 1 get connected to DB instance 2 automatically.

so Please give me the solution for this is it possible or not if yes please share me how to do this.

Accepted Solutions (0)

Answers (2)

Answers (2)

luisdarui
Advisor
Advisor
0 Kudos

Go with AlwaysOn for the database:

https://blogs.msdn.microsoft.com/saponsqlserver/2015/10/23/sql-server-alwayson-summary/

And for the SAP, if you want high-availability you need clustering. For this you would need a SAN drive. For SQL Server, with AlwaysOn Availability Groups, you don't need shared storage, each database server has its own storage.

Bets Regards,

Luis Darui

Sriram2009
Active Contributor
0 Kudos

Hi Sagar.

1. You have to refer the SAP HA installation guide refer the links

https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=7&cad=rja&uact=8&ved=0ahUKEwjz1fPxqrDL...

Refer the SAP guide - SAP NetWeaver Guide Finder

2.   Refer the snapshot of  Windows Failover Cluster on MS Sql DB.

If any one of the node failed other node will take care the SAP application & DB instance. SAP supports Two nodes in Cluster environment.

Regards

SS

Former Member
0 Kudos

Hi Sriram,

But i want to understand that if i don't have share drive(san storage) then.

It is os level storage means i have 2 different os for data base and 2 different os for sap Client so is it possible to implement HA.

Sriram2009
Active Contributor
0 Kudos

Hi Sagar.

1.  The way you are asking is possible but it will have lot of manual work will be there you may have more than 30 to 60 min down time will be there.

2. In Windows Failover cluster mention with in a second time(one requested time out) switch the system from one to another node your operation will not stop for an hour time.

Regards

SS