TI proscess from one view to another is slow

[kastellu]

Him

I'm ussing TM1 9.5.2 version. Does anyone know why there are problems with copying data using TI from one view to another in the same cube. The process just stops and after a while (several minutes) slowly goes on.

Regards

[Christopher Kernahan]

Hi kastellu,

I suspect the issue is less with the version of TM1 you are on than with the process itself. When you say the process stops, and then goes on; what is indicating to you that it has stopped? Have you watched the process with TM1 Top to see what is happening? There may be locking issues occurring when this TI is running.

You can you also try putting a PROCESSQUIT in the TI to see where it slows down, moving it further along if all is ok up to that point.

[tomok]

Your slowdown/locking issues is probably due to locks placed on the cube. As a general rule, I never read and write to the same cube in one TI process to avoid problems like this. if I need to copy one version to another in the same cube I always split the process in two, extracting to a flat file in process 1 and loading the flat file back in to the cube in process 2. Give it a try, I think you'll find it will actually run faster than one process.

[paulsimon]

I would also look at your view. If your source view does not have skip consolidations and skip rule calculated values then it is liable to be slow. It is usually faster to extract at base level and then use an additive load to summarise if necessary

Regards

Paul Simon

[Martin Ryan]

Also take a look at this thread. http://www.tm1forum.com/viewtopic.php?f=3&t=5132 If your cube has a lot of feeders (and rules) then this can significantly impact write back speed as feeders get fired when a value is written to the cube.

Martin