![]() | |
![]() |
| | Thread Tools | Display Modes |
#1
| |||
| |||
|
#2
| |||
| |||
|
|
We have three MOLAP cubes in our environment and lately with no apparent changes to the Cubes or any other parameters, the cube processing time has quadrupled. The server is shared by SQL server and Analysis Services and with a processing time change from 5+ hrs to 20+ hrs, the cubes have become useless as it is time to reprocess them all over again by the time the previoud run has completed. I am not familiar with the internals of how the cubes work but know that only one of them has gone berserk. Being a DBA, I reindexed the cubes database, moved the processing folder to a different location to avoid any disk issues, increased allocated processing memory, rebooted the server, but with no improvement. Well, actually it was back to normal for one day after the reboot and is now back to normal. To make this even worse, we have the same scenario happening on the production and reporting servers. The intention was to eventually shut down the production runs once the validation that the migration to reporting was good thru data validation. Any suggestions are appreciated. Thanks. |
#3
| |||
| |||
|
|
We have three MOLAP cubes in our environment and lately with no apparent changes to the Cubes or any other parameters, the cube processing time has quadrupled. The server is shared by SQL server and Analysis Services and with a processing time change from 5+ hrs to 20+ hrs, the cubes have become useless as it is time to reprocess them all over again by the time the previoud run has completed. I am not familiar with the internals of how the cubes work but know that only one of them has gone berserk. Being a DBA, I reindexed the cubes database, moved the processing folder to a different location to avoid any disk issues, increased allocated processing memory, rebooted the server, but with no improvement. Well, actually it was back to normal for one day after the reboot and is now back to normal. To make this even worse, we have the same scenario happening on the production and reporting servers. The intention was to eventually shut down the production runs once the validation that the migration to reporting was good thru data validation. Any suggestions are appreciated. Thanks. |
![]() |
| Thread Tools | |
| Display Modes | |
| |