burnstripe wrote: Tue Feb 02, 2021 7:59 pm When you trigger stopping the service the model attempts to save all the data in memory to disk (aka a savedataall). Before the service can commit this data to disk any locks/running processes will be resolved first before coming to a halt. If it's taking a long while to stop they'll be something locking or running.

By ending via task manager you're effectively terminating this process so when the service reboots it'll only contain data from when the last savedata occurred. Any transactions newer than this it'll attempt to recover from the transaction log but it may be unsuccessful and could take a while.

1st thing to check when it occurs again though is look at what threads are running on the server, tm1top, operations console and paw admin monitoring will all show what threads/activity is running. Tm1top.exe is located in the bin folder of your install on the server. You may need to configure a file Tm1top.ini to get it working.

If that doesn't help show who and what process/action is causing the lock then it may be the tm1s.log has become corrupt and it is hanging trying to restore it. In which case stop the service take a back up of the file and delete the original tm1s.log and then restart the service.

There's a lot of potential scenarios that could cause this behaviour and it may even be specific to your version of the software. But looking at active threads is definitely the next step I would take