Replication failes in Distributed BO system

[mahieuk]

Hello

we have a distributed system BI 4.0 SP7 platform existing out of a Web Tier (TomcatServer) and 2 BO processing servers
(ProdBOServer and FailOverBOServer).

We are trying to replicate the BO objects from ProdBOServer (origin) towards the FailOverBOServer (destination), but that doesn't seem to work.

We have created a simple object (webi report) on the ProdBOServer and added a replication list with this object.

On the FailOverBOServer, we created a remote connection with the following settings:

     Remote System Web Service URI: http://TomcatServer:8080/dswsbobje

     Remote System CMS: ProdBOServer

     Username: Administrator

     Password: xxxx

Then we created a new replication job based on the replication list of the ProdBOServer.

After running the replication job, the status is success but no object has been replicated.

When looking at log file, we got a error that our browser didn't support xml 1.1. When displaying the log with a different browser, we get

We have also put up a TraceLog (high) to the AJS with the following result:

...

|1528|0|4|0|CMC.WebApp|TomCatServer:2992:74.16085:1|CMS.runJobs|localhost:3208:3464.2947714:1|.run|FailOverBOServer:9044:28.8:1|CqBhSBmWS0NPjGTuOu4ZmCg3efd|||||||||com.businessobjects.js.child.federation.ReplicationProcessingSubsystem||Failed
to find EnterpriseNode for server with friendly name 'BI1.AdaptiveJobServer'

...

SDK error occurred at the origin site.

cause:com.crystaldecisions.sdk.exception.SDKServerException: All the
servers with CMS ProdBOServer:6400, cluster
@ProdBOServer:6400, kind cms which host service null, are down
or disabled

cause:com.crystaldecisions.enterprise.ocaframework.idl.OCA.oca_abuse:
IDL:img.seagatesoftware.com/OCA/oca_abuse:3.2

detail:All the servers with CMS ProdBOServer:6400,
cluster @ProdBOServer:6400, kind cms which host service null,
are down or disabled

detail:SDK error occurred at the origin site. All the servers with CMS
ProdBOServer:6400, cluster @ProdBOServer:6400,
kind cms which host service null, are down or disabled

Can anyone help us out?

Thx.

[denis_konovalov]

Are both your source and destination servers part of the same cluster ?
The picture suggests that they are...

[mahieuk]

Yes, they are

[denis_konovalov]

in that case why are you using federation ?
If they are part of same BO cluster - they already share the same CMS DB/FRS or you have to replicate CMS DB/FRS outside of BOBJ.

[mahieuk]

My mistake, 2 separate CMS systems with their own cluster

[Former Member]

Sorry for the late reply to this, from the XML output you showed above there is an "ERROR" "NULL".

There are a few other logs you should check.

1. Replication works by connecting through the dswsbobje web service and requesting the objects, etc. It's possible that you'll get a better error from there.

2. Use the end 2 end tool to schedule the test job until you figure out what the problem is. This way, you can deactivate the logs - set then to unspecified and the SAP Passport which is sent from the scheduled job will turn on the logs on all the relevant servers in the workflow.

Since I believe you have an incident open as well - please include those logs in the incident and we'll work with you. We can potentially post the answer back to the forums so that this post helps others in the future.

--

Kind regards,

Patrick Bulteel

[mahieuk]

Hello Patrick

can you tell me what that end 2 end tool is?

Thanks in advance

Kind regards

Kristof Mahieu

[Former Member]

Hi,

Sorry - I should have included the link...

The KBA for this is  1861180 (link below.)

http://tiny.cc/w88zqx

However, if as Denis says, they are part of the same cluster - this means that they are using the same DB/FRS and hence you would see the two nodes in the CMC servers list - then there is no reason to do what you're asking to do.

You basically are "replicating" to the same system.

However, that was not my original understanding of the issue... So, just let us know and we can continue.

-P

[Former Member]

I am not able to figure out what you mean by replication of webi report objects. Do you mean that you are promoting the objects? Please elaborate.

[Former Member]

@Arvind,

Mahieu is talking about using Federation, created replication jobs to replicate source and destination.

[mahieuk]

@Arvind,

Indeed as Jawahar mentioned, the goal is to replicate all repository objects from source to target through scheduling. But by eliminating the change that overload causes the issue, we started with a simple report. We have done the same test with other objects (access levels, ...). Each time with the same error.

[Former Member]

Thanks for the inputs Jawahar and Mahieu. Guess need to go through federator closely. haven't done much on this will need to go through this.