Duplicate check
Before submitting a bug report, I have:
What is the behavior?
I had a download on a private tracker fail to start due to the tracker being down, but it wasn't getting strikes. In the logs I can see the message:
skip | multiple StallRule rules matched | [file]
Job QueueCleaner
Instance [instance]
Job Run 019d4d9e-1920-78fa-8c3b-e863dc8898d8
(These are the only logs for this job run - I didn't actually get to enable verbose logging since the download was cleaned once I disabled a rule)
The download was from a torrent file rather than a magnet link, which I assume is why it didn't count as stuck fetching metadata?
I wanted to configure the rules so that:
- public tracker downloads at any progress can be deleted
- private tracker downloads at 0% (they haven't started, like in this case) can be deleted
- private tracker downloads >0% (they have started) can't be deleted
I attempted to configure this with 3 rules:
- Public, 0-100%
- Private, 0-0%, Delete private from client enabled
- Private, 0-100%, Delete private from client disabled
In this case, the download matches rules 2 and 3, triggering the multiple match skip.
If I set rule 3 to 1-100%, it warns there's a gap 0-1% for private. Incidentally it doesn't warn that there's overlapping rules until it appears in the logs, it might be useful if the settings page warned about this like it does for gaps.
Am I missing a way of configuring the desired behaviour? Or is there something wrong with my planned rules, which is why it's not supported?
A possible fix for this would be to make the lower bound exclusive (so 50-100% is actually >50%, ≤100%), and add a checkbox to match downloads that haven't started (progress is exactly 0%). I think that changing the lower bound to be exclusive would have minimal impact on existing rules, since it's unlikely that download progress will ever fall exactly on the threshold.
Which operating system do you use?
Linux
What type of deployment do you use?
Docker container
Relevant log output
Anything else?
No response
Duplicate check
Before submitting a bug report, I have:
What is the behavior?
I had a download on a private tracker fail to start due to the tracker being down, but it wasn't getting strikes. In the logs I can see the message:
(These are the only logs for this job run - I didn't actually get to enable verbose logging since the download was cleaned once I disabled a rule)
The download was from a torrent file rather than a magnet link, which I assume is why it didn't count as stuck fetching metadata?
I wanted to configure the rules so that:
I attempted to configure this with 3 rules:
In this case, the download matches rules 2 and 3, triggering the multiple match skip.
If I set rule 3 to 1-100%, it warns there's a gap 0-1% for private. Incidentally it doesn't warn that there's overlapping rules until it appears in the logs, it might be useful if the settings page warned about this like it does for gaps.
Am I missing a way of configuring the desired behaviour? Or is there something wrong with my planned rules, which is why it's not supported?
A possible fix for this would be to make the lower bound exclusive (so 50-100% is actually >50%, ≤100%), and add a checkbox to match downloads that haven't started (progress is exactly 0%). I think that changing the lower bound to be exclusive would have minimal impact on existing rules, since it's unlikely that download progress will ever fall exactly on the threshold.
Which operating system do you use?
Linux
What type of deployment do you use?
Docker container
Relevant log output
Anything else?
No response