I have an issue and was hoping if others have also encountered a similar issue
Data loss on one particular cube after a Windows Service Restart
TM1 Version 2.0.5
Users enters data in about 5 cubes, however data is lost on only ONE cube after a service restart.
Data is entered using Contributor by the users and visible when I check it using Architect.
Data entered directly in Architect is unaffected.
The transaction log looks good and shows all the values that have been entered.
Last month, We lost about 10 days of data on that particular cube which I could fully restore using a TI process and the transaction log.
Before taking a backup, I right click on the model and do a 'save data'
Changes suggested by IBM:
In tm1s.cfg, One Configuration change made, MTQ=0, multi-threaded optimization is turned off.
Execute a TI process to "save" data (SaveDataAll in epilog) before stopping and starting service.
This has helped more or less fix the issue. 9 out of 10 tests done were successful, but I am still a bit skeptical
and think that there is something else that is going on.
Another interesting aspect is that after implementing IBM changes data loss happened only when the old value is non-zero.
So, if the old value is say 100 and I change it to 200, it saves it
if the old value is zero and I change it to 50, it is not saved
Another thing I noticed is that I (Adimin) encounter the problem much lesser than normal users.
Any advice or suggestion would be highly appreciated.
Thanks in advance.
1 post • Page 1 of 1