TM1 Forum

Hi gurus

Im trying to run a process that updates Cube A. At the same time users are reading, using Cognos 10 BI, from Cube B and since there are No rules or feeders between these cubes I thought i could run the process without it affecting users reading from Cube B. What happens is that anything running (reports analysis etc) based on Cube B will just wait for the job to finish on Cube A. As soon as the job finishes the users can work with Cube B again.

Both cubes are on the same TM1 server.

I checked the ParallellInteraction in tm1s.cfg and its set to true so that should not be the issue.

Anyone got an idea here?

/P

Are you doing any meta changes within the process? By that I mean dimension or subset updates? Is the process running within a chore that also has meta updates within it?

Hi

No Im only loading data to cube A. The dimensions are shared between the two cubes though, which I missed to inform.

/P

I wonder if it has something to do with BI loading all it's meta data before displaying the view? Have you tried loading cube A while somebody views cube b in perspectives?

Have not tried that but since it soes not even work to browse the cubes dimensions in BI when the process is running.

/P

I vaguely remember something about the way logging was set causing issues with locking.

At one point during testing we had to actually create a dummy if statement around the generated statement that included the CUBESETLOGCHANGES. Basically if(1>2) kind of thing. Then you either have to write to the logging cube directly through a cellput.

You could try seeing if you can avoid the CUBESETLOGCHANGES command and test. Not sure if this is still an issue but might be worth a try.