Closed JakeWharton closed 10 years ago
You can see the memory drain to 0%, a quick bout of GC thrash, and then the ultimate OOM. Maybe batching or doing background processing would fix.
Ah, ok. It's probably a bit too aggressive with the number of threads it creates when retrieving the movie data.
How many files do you have?
Not sure what extensions it looks for, but here's .mkv
s:
jake@theflame:/tanker/media$ find . -name *.mkv | wc -l
4140
Ok. That's quite a bit more files than what I have. It looks like it's running out of memory while building the initial list of files. It looks for files with the following extensions: ".3gp", ".aaf.", "mp4", ".ts", ".webm", ".m4v", ".mkv", ".divx", ".xvid", ".rec", ".avi", ".flv", ".f4v", ".moi", ".mpeg", ".mpg", ".ogv", ".rm", ".rmvb", ".mov", ".wmv", ".iso", ".vob", ".ifo", ".wtv", ".pyv", ".ogm", ".img"
I got to the API call stage using a single subdir which only has 515 files in it.
Ah, ok. So you ran into the OOM with a directory with only 515 files? I'll try running it from the root directory on my NAS.
The OOM was on the 4000+ file directory. It never got the the stage where it makes API calls to TMDb. When I ran it on a subdir with only 515 files it got to the stage where it's calling TMDb for data (and still is...).
I haven't dug into the code yet, but I get OOMs when trying to synchronize my entire share.