If I understand correctly, you want to make sure that you catch the correct logs with a certain logging threshold.
There is a difference between the log threshold (where you set if for instance only errors are displayed or also warnings or all information) which you have to be very careful about because it gets very verbose, on the one side. On the other side, you have the serverlog.txt and this is what I THINK you intend to find.
For the serverlog.txt there are two parameters:
The first tells PPM how large a serverlog.txt may become before actually saving and archiving this (to the same folder as the one that is currently written to). Default is 250k and you can increase this to the level that you need. The more verbose your logging is, the quicker this file is written to and fills up.
The second parameter determines, how MANY files are created as archives until the oldest one is overwritten. Default value is 20.
ROTATE_LOG_SIZE = 1024K
MAX_BACKUP_INDEX = 20
The active log is written to until it has reached 1 MB in size. It is then archived, a new file is started. this continues until all 20 backup files are written full, then the oldest one is overwritten. this means that you can have a log-folder with 21 log-files (1 active, 20 backup). Calculate carefully how much space you have prior to setting this up.
Hope this helps, else let me know what info is missing. Incidentally, this is information from the admin guide pages 465 foward.
hmm, I remember that there was an issue with 8.03 rolling, meaning that no backup files were created and instead, the active one is overwritten immediately. This happened when the windows service was used to start PPM and not kStart. Now, a 5k serverlog won't help you much, I advise to set to 250k. In addition to that, I would test a start with kStart.sh instead of the windows service just to test that behaviour. I just checked and it seems that the defect I reported in 8.0x (and solved in 8.03) seems to be present also for 9.1x and is planned to be solved in 9.12 ...
Check QCCR1L28261 on SSO.
Thus, if you start PPM via kStart.sh, the problem should not occur and backup-files should be created.
I'm not quite sure what you refer to when you mean 'when I disconnect from my admin session' but generally, the answer is that we have a special user that we install PPM with that is not admin of the box we connect to. That way for instance I can disconnect from the RDC-session when logged in with my win admin account and the PPM owner-based cygwin box stays open.
Maybe someone else has an opinion on how to solve this issue but if you use RDC, then I would use a different account.
It may take a minute or two after you change logging.conf, but PPM will see that the file has changed and reprocess it. When this happens, you will see a message in the log file reporting that "Logging reconfiguration complete" or something like that. PPM does not need to be restarted.