on 01-17-2008 6:55 AM
Hi,
I have CI - 02 on one host and dialog instance -03 on another host.
I have created a logon group by name public and assigned these 2 instances.
My question is if one instance got down how the user will automatically switch over.
It's cluster environment already installed.
Let me know if u need information.
Regards,
Saravanan.S
Hi,
If I undestand correctly, you have a CI on cluster (MSCS ?) and a SA on a third host.
Your logon group includes both the CI and the SA.
It means that your users are, eithe connected on the CI or the SA.
With this architecture, you don't have High Availability.
If you user is connected on the CI and there is a cluster fail over, the user will be disconnected if the CI node failed.
If your user is conencted on the SA and there is a SA failure, the user is disconncted.
For HA, it's better to have at least 2 SA on 2 different hosts and to include only the 2 SA in the logon group.
Regards,
Olivier
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Hi Olivier,
Sorry for the late resposnse.
Went out for a vaction.
Now I have installed 2 DI outside of MSCS.
define logon groups for the 2 DI as public.
when i login under the group, and shut down one di the users are not automatically switch over.
I tried ensmon.exe and got replication is active as message.
Let me know how to proceed further.
Regards,
Saravanan.S
Hi Saravanan,
>when i login under the group, and shut down one di the users are not automatically switch over.
This is nominal. If you stop one DI, all the users conencted to it are disconnected but they can reconnect immediately to one of the remaining DI.
Don't believe that High availability is something magical !
Regards,
Olivier
Saravanan,
If you need the impossible then you are in trouble !
The purpose of a cluster and several DI is that, in case of failure, users are able to reconnect almost immediately.
For RFC calls the service interruption is virtually invisible.
The availability of SAP running on a cluster is higher than without a cluster, that is why it is called a
high availability solution and not an absolute availability...
Regards,
Olivier
User | Count |
---|---|
81 | |
10 | |
10 | |
9 | |
7 | |
6 | |
6 | |
5 | |
4 | |
4 |
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.