TM1 Forum

Hi All,
Currently TM1 APP Instance takes 2 hours and 45 mins to come up after services are restarted. Data folder is approx. 170 Gigs and server holds models like Profitability etc. approx. 80 - 85 Gigs (Profitability model developed on TI processes very very minimal TM1 Rules usage)
Can any body suggest ways to improve performance of application

Few details
TM1 APP Instance hosted on IBM Cloud on Bare metal Server
RAM - 256 Gigs, CPU 48 Chores
PersistentFeeder=T in tm1s.cfg

searched for old threads for clues but did not find may be i missed them, please help me if some body found them.

Thank you

ttesla wrote:Currently TM1 APP Instance takes 2 hours and 45 mins to come up after services are restarted. Data folder is approx. 170 Gigs and server holds models like Profitability etc. approx. 80 - 85 Gigs (Profitability model developed on TI processes very very minimal TM1 Rules usage)
Can any body suggest ways to improve performance of application

Few details
TM1 APP Instance hosted on IBM Cloud on Bare metal Server
RAM - 256 Gigs, CPU 48 Chores
PersistentFeeder=T in tm1s.cfg

searched for old threads for clues but did not find may be i missed them, please help me if some body found them.

Moved to the correct forum. Please don't post TM1-specific issues in General.

You've indicated that you have persistent feeders turned on but is the server actually using them? Specifically, does the server log (not the transaction log) report any issues with the feeder files which are causing them to be deleted and needing to be re-evaluated?

Have you tuned on Performance Monitor (make sure that you turn logging off for the Performance Monitor cubes) and checked to see whether there are any values that are completely out of whack with what you would expect? (Something that might indicate overfeeding on a massive scale, for instance.) There are a bunch of other things that I can think of (issues with disk speed, for instance) but best start with the obvious ones.