CM Repository Garbage Collector - System ID, which one to pick?

[Former Member]

I followed instructions in <a href="http://help.sap.com/saphelp_nw70/helpdata/en/92/99643e67724644e10000000a114084/frameset.htm">http://help.sap.com/saphelp_nw70/helpdata/en/92/99643e67724644e10000000a114084/frameset.htm</a>.

When I edit the Task, then I can select a CM System ID with 8 numbers e.g. 12345678. Where does this number come from? Is this the system that I need to select?

When I click on "See also: Content Management System" and edit the System, then this system consists of a number and a description.

I somehow miss the relations. This is quite confusing. Do I need to create a new system, or is the system provided sufficient?

Thanks,

Dieter.

[frank_friedrich]

Hi Dieter,

it is indeed a little bit complicated.

So you have a Portal installation and normally one dispatcher and one java server. This java server has one ID. You can add more java servers (like using the Config Tool). This second java server gets as well an ID but of cause a different one. This two ID's are as well listed in KM as Content Management Systems and this IDs are exactly the same as the ID's you see in the Config Tool or the Visual Administrator Tool for the Java Server Nodes. Now in KM you have to assign for different global services on which server node this tasks should run. So you must assign this tasks to exactly one CM System.

Best Regards

Frank

[Former Member]

Hi Frank,

thanks for great answer! So if only one number is presented, then I just select that number, correct?

Does it make sense that only one number is presented? I always have a server and a dispathcer, don't I?

Thanks,

Dieter.

[former_member189326]

If you are not in a clustered environment (and it sounds like you are not, since only one node ID is being presented), then this step is not necessary. Only if you have multiple nodes on which tasks could execute do you need to bind tasks to a specific node. This is because there should only ever be one instance of a task, and it allows you to configure your server landscape for optimum load distribution. So for you, it's not complicated.

[frank_friedrich]

Hi Dieter,

Boris answer is correct.

If you have a normal installation (a simple once) then you have on dispatcher and one java server. The only task for the dispatcher is to dispatch the requests to the available servers. That's all. So the jobs (tasks, processes, threads, ...) are running at the java server. That's why you have to assign a server node. Never ja dispatcher and no dispatcher will be found in the list of content management systems. If you have only one java server you doesn't have to do anything. When you have two or more than you have to assign a java server node.

But take care when removing a java server node when you have assigned tasks to this java server node.

Best Regards

Frank