TI restarts at some point

[Nick_Blommaerts]

This may be a bit abstract, but here goes:

Basically, I have a master TI that launches a bunch of other TI's in a certain order. When doing this on my development environment, all is well and the full set of TI's finish in 5 minutes. An exact copy of this TM1 server, migrated to my test environment, ends up cutting the string of TI's at some point and restarts. This behaviour is repeated for hours if I let it run. The TM1 servers are an exact copy of each other, their behaviour is just different when ran on different servers. The problem occurs when running the TI's, which extract data from several cubes and summarize them in one reporting cube, but only for the production plant is continues to loop for a long time (it has never finished successfully so far). The server on which it fails is more powerful though, so it seems unlikely that it would play a role.

Has anyone ever experienced such behaviour? All I could find was one thread from a couple of years ago, but there was no real conclusion as to what may be the issue.

[Alan Kirk]

This may be a bit abstract, but here goes:

Basically, I have a master TI that launches a bunch of other TI's in a certain order. When doing this on my development environment, all is well and the full set of TI's finish in 5 minutes. An exact copy of this TM1 server, migrated to my test environment, ends up cutting the string of TI's at some point and restarts. This behaviour is repeated for hours if I let it run. The TM1 servers are an exact copy of each other, their behaviour is just different when ran on different servers. The problem occurs when running the TI's, which extract data from several cubes and summarize them in one reporting cube, but only for the production plant is continues to loop for a long time (it has never finished successfully so far). The server on which it fails is more powerful though, so it seems unlikely that it would play a role.

Has anyone ever experienced such behaviour? All I could find was one thread from a couple of years ago, but there was no real conclusion as to what may be the issue.

Are you sure that there aren't any other processes running on the production server which could cause locking contention and therefore a rollback?

[Nick_Blommaerts]

This may be a bit abstract, but here goes:

Basically, I have a master TI that launches a bunch of other TI's in a certain order. When doing this on my development environment, all is well and the full set of TI's finish in 5 minutes. An exact copy of this TM1 server, migrated to my test environment, ends up cutting the string of TI's at some point and restarts. This behaviour is repeated for hours if I let it run. The TM1 servers are an exact copy of each other, their behaviour is just different when ran on different servers. The problem occurs when running the TI's, which extract data from several cubes and summarize them in one reporting cube, but only for the production plant is continues to loop for a long time (it has never finished successfully so far). The server on which it fails is more powerful though, so it seems unlikely that it would play a role.

Has anyone ever experienced such behaviour? All I could find was one thread from a couple of years ago, but there was no real conclusion as to what may be the issue.

Are you sure that there aren't any other processes running on the production server which could cause locking contention and therefore a rollback?

I'm sure that this was not the case. I've been looking for this issue since Friday, obviously as soon as I post for help here it appears I have found the issue.

On my server for this country, it appears someone has added the following at some point to the config file, which I completely overlooked:
PerformanceMonitorOn=T

After removal, the loops are gone. I've tried to put it back and the loops are back again as well. What I find strange is that on one server it doesn't give issues, on the other it does, even though the specs of the servers are largely the same (see attachment).

[Gregor Koch]

If the removal of
PerformanceMonitorOn=T
fixed your issue, you might want to check whether the cube logging for the }stats cubes is turned on and turn it off. Not sure if the logging of those might actually have caused your problem but it is still recommendable to do turn it off just in case you want to use Performance Monitor.