Zero byte corrupt file causes excessive IMDB metadata downloads

Request Type:

MCEBuddy Version and Type (32bit or 64bit):
2.5.4 64bit

Operating System and Type (32bit or 64bit):
Windows 10 64bit

Summary of the problem or suggestion:
Due to a disk space issue, my NextPVR recordings were failing and leaving zero byte files. MCEBuddy tried to process these files and failed after retrieving metadata from IMDB. The engine retried processing these files continuously resulting in a very high amount of data downloaded from IMDB. I was alerted to the issue when I was notified by my ISP that I hit my data cap. Upon investigation the application was pulling down over 100 GB of data from every day as it was trying to process a handful of corrupt files. There needs to be some sort of back-off or circuit breaker introduced when retrying files that fail due to zero file size or prevent retries on files that are unlikely to recover.

Steps to replicate the bug:
Attempt to process an empty TS file

Screenshots: (1.2 MB) (8.8 MB)

Thanks for reporting this. I’ve been able to replicate it. Technically, it’s working as expected since the file has valid metadata in it’s title so it’s trying to match it on IMDB, however since this series has over 1000 episodes it takes times and traffic. However since your file was 0 bytes it’s a moot exercise.

We’ve optimized this further in today’s 2.5.7 BETA build. Now it’ll fail any files which are 0 bytes right the beginning and log the error accordingly. We’ve also optimize the IMDB matching to make it faster and reduce the time/traffic.

1 Like