Page 1 of 1
Feeder computing time
Posted: Thu Aug 07, 2014 2:34 pm
by deepakjain2020
Hi All,
General question again
Can many conditional feeders in a cube lead to increase in feeder computation time?
Regards,
Deepak Jain
Re: Feeder computing time
Posted: Thu Aug 07, 2014 7:13 pm
by declanr
deepakjain2020 wrote:Hi All,
General question again
Can many conditional feeders in a cube lead to increase in feeder computation time?
Regards,
Deepak Jain
There are many ifs and buts when it comes to the detail of this but the generic answer to the generic question is that it will. Generally speaking adding more feeders (conditional or otherwise) will of course lead to more feeders having to be processed; therefore meaning it would take longer to do so.
Re: Feeder computing time
Posted: Fri Aug 08, 2014 5:37 am
by deepakjain2020
Thanks Declanr
You are right, I have removed conditional feeders from rules and had check in TM1server log.
The feeder computation time has changed and it is consuming less time than earlier
Regards,
Deepak Jain
Re: Feeder computing time
Posted: Fri Aug 08, 2014 8:04 am
by David Usherwood
The conditionals would have been set up to reduce server size. How much larger is your server's RAM footprint now?
Re: Feeder computing time
Posted: Fri Aug 08, 2014 8:29 am
by deepakjain2020
David Usherwood wrote:The conditionals would have been set up to reduce server size. How much larger is your server's RAM footprint now?
Hi David,
There was a overfeeding of cube.
RAM size - 24 GB
Before making changes to feeders it was consuming around - 14 GB (Including TM1 data) and now it is around 9.5 GB.
Later added conditional feeders, which was a good idea but resulted in increase of feeder computation time which indeed impacted report response time also. So replaced with respective elements rather than having conditional feeders.
Regards,
Deepak Jain