TM1 Forum

Hi,
I have a cube called cube A that is heavily loaded with rule and feeder. When I load the data (Forecast1) from *.csv via TI to the cube, the memory will keep increasing. Normally, the loading will took bout half an hour to finish.

If I load the same set of data to another version called Forecast2 using the same TI, it will loop infinitely and cause the server memory hit the max. Hence, the cube A always only contain one version of data.

Fyi, the saving time for the cube A's rule is taking bout half an hour. And, the file for loading is bout 200 records.

I have created another cube - Cube B. This cube has same dimensions as Cube A. But, no rule in this cube B. The data loaded successfully for both versions and it is quicker.

I'm thinking to feed the data from Cube B to Cube A instead. But, would this cause the increase of memory since need to feed everything from cube B to cube A? Or, are there any other solution?

Would appreciate your help on the above. Thanks.

Sigh, we seriously have to look at putting some restrictions on posting into the Bugs forum...

Please read the How to make a bug post thread on what is and is not a bug.

Thread moved to the TM1 forum.

If 200 records take half an hour to load, your feeders are probably excessive. Trace them through and decide whether you need them.

David Usherwood wrote:If 200 records take half an hour to load, your feeders are probably excessive. Trace them through and decide whether you need them.

Given that you can reasonably expect to process 10K to 20K records/second when loading from a CSV that is a slowdown in the order of 10^5. Excessive may be under doing it a tad, I'd say outrageously excessive. Get someone (who knows what they are doing) to review the rules and feeders associated with the cube that the data is being loaded into.