I have Sonarr properly pointed to Jackett which points to IPT. This is tested working as it pulls most other shows, but it is not properly pulling Jeopardy. It appears that Sonarr looks for Jeopardy by year and episode number (e.g. 2019x110) while IPT and pretty much every other torrent site I’m finding have it listed by air date since there are so many episodes (e.g. 2019 06 05). I’m assuming this mismatch is why Sonarr is not pulling the episodes.
For reasons I don’t understand, Wheel of Fortune seems to be fine. Same mismatch exists though.
I mistakenly posted this to GitHub as a bug originally and got a response that I needed to set the Series type to “Daily” - I’ve done that, but still no luck. Not sure if I need to force a refresh or what, but it still won’t pull Jeopardy episodes from IPT.
Update: If I click on the episode title under the “Wanted” tab, go to the “Search” tab and hit “Automatic Search,” it works. It will pull a torrent and put it under “Activity” and pass it off to Transmission for download.
But if I click “Search All Missing” it runs through the list very quickly and comes back empty-handed.
Search all missing (and a season level search) uses season searches, which IIRC for daily is the year + month in a specific format. Jackett likely doesn’t understand that format and doesn’t pass a request to the indexer that it returns results for.
((Debug logs)) will show the exact request Sonarr is making, Jackett may need to be updated to support that.
So I enabled debug logs and then did both - the “search all missing” as well as the specific episode -> “automatic search” force - and these are the logs that were produced showing the unsuccessful request for the first attempt, and the successful request for the second. I’m not really sure what I’m looking at here so maybe you can tell me what the difference in the requests is? Wouldn’t “search all missing” effectively just be making the same request the “automatic search” does, but just multiplied by how many things are in the Wanted queue?
No, it groups by season to reduce the number of API calls it makes.
No month, but the year. Which probably results in Jackett making a request for Jeopardy S2019 as the tracker itself doesn’t have indexing to support Jeopardy + 2019.
Okay, so that makes sense, but now it’s been 24 hours and a new episode of Jeopardy just streeted. Sonarr knew about it, because it came up in the Wanted list with an Air date of 12 hours ago (today it says “Yesterday”). But it still failed to queue anything up for download until I click the specific episode and do the “automatic search” again. So it’s not really performing in the automated capacity I am hoping for.
Is whatever cron that runs to get new missing items just doing the same API call that “search all missing” does, thus not giving Jackett anything to work with? Is there a way to fix this? Clear out the older episodes or something so that it will feel more comfortable only asking for one thing at a time?
No, it doesn’t search, as explained in the ((FAQ)).
You’re going to need to look at ((debug logs)) to see if Sonarr ever sees the release on the RSS feed and why it rejected it. Make sure ((debug logging)) is enabled and check the logs after the next release comes out.
I still have debug mode on, and did skim the logs before posting… didn’t see Jeopardy even mentioned except for where I forced the automatic search. So perhaps I’m not understanding then… is Sonarr only capable of finding things automatically that are advertised on the RSS feed, but can use Jackett to run an active search against the tracker’s entire database?
Could this mean that even though the tracker carries the latest Jeopardy episodes, they just don’t advertise them on their RSS?
That is how Sonarr finds new releases, it only searches automatically to find a replacement for a failed download.
It can do that for any indexer that supports searching, including Jackett.
RSS is the latest releases, I’ve never known an indexer to not include specific releases. Beyond that it could be too long of an RSS Sync interval (the default of 15 should be fine), using Jackett’s aggregate indexer (/all instead of each indexer individually in Sonarr meaning you can get more gaps in the RSS Sync/less search results as Sonarr will usually avoid paging).
Jackett should only be contacting one tracker, IPT, since that’s all I have configured. Not sure if /all would still limit the results of the only tracker it contacts. The only reason I use Jackett is because it appears Sonarr was not able to integrate with IPT, because even using it’s RSS, it wasn’t getting anything, even stuff that I knew was being advertised on the RSS. I was told on other forums that Jackett was a requirement to properly use private trackers like IPT, so that’s really all it’s doing. It logs into IPT on my behalf and presents a Torznab feed for Sonarr’s use.
Right, I know how Jackett operates. You can use IPT directly with Sonarr, but that only works for RSS, not searching (since they have an RSS feed, but don’t have an API).
If the debug logs don’t show Jeopardy before you went and searched it then Sonarr had no opportunity to grab it, also look for Indexer "indexer name" rss sync didn't cover the period between X and Y UTC. Search may be required. that is Sonarr complaining that the RSS feed from the indexer got more items than Sonarr could request (as well as more than it could page back to request), which is max 1000 (10 pages of 100 items), but could be 100 if IPT doesn’t allow paging, in which case you’d want to shorten the RSS Sync interval.