I’ve run across this in a few implementations now. You go to run the OCSLogger on an older Lync server and there seems to be way too much data in there, old data in fact. You can see in the below image, I ran the logger for literally 2 seconds, but somehow have 9 SIPStack files. This can happen when trace files refuse to be deleted or overwritten.
The quick answer is to close your logger, find your tracing folder (often C:\Windows\Tracing) and move the old ETL files, or all the files in the directory somewhere else.
You should now be able to start the logger fresh.