So here’s what I did:
- I quit Tautulli.
- Started Tautulli again and “connected” on the websocket testing website.
- Started watching a show on my remote computer; one other user was already watching a show.
- Let it run for 10 minutes, then downloaded the logs.
Comparing the two logs is difficult, as what I get from the website contains no timestamps. That being said, I knew exactly what time I re-started Tautulli, so that wasn’t a big deal.
What I found was that after I restarted Tautulli, the ten minutes’ worth of logs were clean: no garbage characters or data. But if I looked at earlier timestamps, I found those non-standard characters in the log. I don’t know if restarting is the solution, because I’ve restarted Tautulli many times during my testing, and the garbage data always shows up again at some point. The logs from the website were also clean.
The big limitation in collecting logs from that website is that it just dumps it to a text field on the page; copying from that field only yields partial log information (about 45 lines, versus 671 generated by Tautulli). Upon further use, later Tautulli logs show more errors.
