Strange Cognos job issue with TM1

[macsir]

Hi, all
I am wondering if you have any similar problem before which can give me some idea.
I have cognos job to refresh reports based on TM1 data. The job can be run in dev and test environment without any issue. In prod, the job sometimes just refreshed some reports and then stopped fetching data from tm1 although in Cognos, the job is still showing executing. Actually it has died already.
The cognos server guy has checked everything on prod and nothing seems to be wrong in configuration.

The cognos is 10.2 and TM1 is 10.1.1. This issue looks like happened randomly. It was fine to run the reports at first several times but after while, this symptom showed and the job was executing forever.
It looks like this issue happened to other databased jobs as well.

Does anyone have any clue? Thanks.

[tomok]

Does anyone have any clue? Thanks.

What kind of analysis have you done on the TM1 Server, i.e., memory situation, checking out the tm1server.log, etc.? Your answer most likely lies there.

[macsir]

What kind of analysis have you done on the TM1 Server, i.e., memory situation, checking out the tm1server.log, etc.? Your answer most likely lies there.

Hi, tomok

There is nothing unusual in TM1 server log and Cognos server log. The cognos guy has made same configuration (memory setting, etc) as the test server but still the issue is there.
I am thinking there is an underlying hardware issue which I can't see.

[kaazimraza]

The cognos is 10.2 and TM1 is 10.1.1. This issue looks like happened randomly. It was fine to run the reports at first several times but after while, this symptom showed and the job was executing forever.
It looks like this issue happened to other databased jobs as well.

Sounds like connectivity issue to me.

[macsir]

Sounds like connectivity issue to me.

But TM1 connection in Cognos is successful.

[kaazimraza]

Sounds like connectivity issue to me.

But TM1 connection in Cognos is successful.

Is the connection between TM1 & Cognos is successful at the times when report fails? And, you mentioned about other database jobs failing as well. Can you give some details on what kind of jobs they are? Like, are they cube build jobs that pull data from underlying database into TM1 cubes, or are they some data movement tasks/ETL within Cognos?

[macsir]

Is the connection between TM1 & Cognos is successful at the times when report fails? And, you mentioned about other database jobs failing as well. Can you give some details on what kind of jobs they are? Like, are they cube build jobs that pull data from underlying database into TM1 cubes, or are they some data movement tasks/ETL within Cognos?

Yes, the connection is always successful. Those database jobs are the ones which refresh reports based on relational database instead of tm1. My job are the one which refresh reports based on TM1.

[macsir]

Just share some experience,
We had submit this to IBM to look after this and they didn't find anything special yet. The issue still exists so far.
But we had some clues. On that server, the two BIBUS services can't get more memory than 2GB for each when it is supposed to get normally. Server Admin Guy confirmed that no issue with the hardware especially the RAM.
This issue doesn't appear on TEST server. That's why TEST is fine so far.

I will keep this updated if any news coz this is the most bizarre issue I have ever seen.