Speed boost on exporting cube into file and importing back to cube
Posted: Thu Mar 02, 2017 5:31 am
Hi all
Just want to share that I have managed to reduce exporting time from 90 minutes down to 20 minutes from a massive rule-based cube. I guess you might know how I did it. Yes, it is runTi, which can max the CPU cores in parallel.
But here is the problem, if you use runTi and executecommand, you lost the control of the process since you don't know when the runTi finishes the export and when to start to import afterwards. The solution is power shell script. You basically wrap everything into the powershell. It has function to monitor all sub-processes it kicks off and then trigger the import process after export. The last piece is to send out e-mail to report that snapshot is done. Nice and Easy!
Just want to share that I have managed to reduce exporting time from 90 minutes down to 20 minutes from a massive rule-based cube. I guess you might know how I did it. Yes, it is runTi, which can max the CPU cores in parallel.
But here is the problem, if you use runTi and executecommand, you lost the control of the process since you don't know when the runTi finishes the export and when to start to import afterwards. The solution is power shell script. You basically wrap everything into the powershell. It has function to monitor all sub-processes it kicks off and then trigger the import process after export. The last piece is to send out e-mail to report that snapshot is done. Nice and Easy!
