Cannot connect to any Indexers

I am having an issue connecting to all of my indexers. I have four right now, and as of 3ish or so weeks ago, none of them connect. I can hit the API outside of Sonarr and it works as expected. But, for some reason, when Sonarr goes to connect, it fails. I haven’t touched the system outside of adding and removing a series. For example, I can hit https://api.dognzb.cr/api?t=caps&apikey=[api key] without issue from a browser (yep, replaced [api key] with my actual api key) and I see the XML

Sonarr version: 2.0.0.5344
OS: Microsoft Windows Server 2012 R2 Standard

Debug Log:

20-9-7 16:55:24.3|Debug|NewznabCapabilitiesProvider|Failed to get newznab api capabilities from http://api.dognzb.cr/

Debug log file: https://www.dropbox.com/s/m58wh7vqoj51z7a/debug.txt?dl=0

Snippet of the issue.

[v2.0.0.5344] System.Net.WebException: The underlying connection was closed: An unexpected error occurred on a send. —> System.IO.IOException: Authentication failed because the remote party has closed the transport stream.
at System.Net.Security.SslState.StartReadFrame(Byte[] buffer, Int32 readBytes, AsyncProtocolRequest asyncRequest)
at System.Net.Security.SslState.StartReceiveBlob(Byte[] buffer, AsyncProtocolRequest asyncRequest)
at System.Net.Security.SslState.CheckCompletionBeforeNextReceive(ProtocolToken message, AsyncProtocolRequest asyncRequest)
at System.Net.Security.SslState.StartSendBlob(Byte[] incoming, Int32 count, AsyncProtocolRequest asyncRequest)
at System.Net.Security.SslState.ForceAuthentication(Boolean receiveFirst, Byte[] buffer, AsyncProtocolRequest asyncRequest)
at System.Net.Security.SslState.ProcessAuthentication(LazyAsyncResult lazyResult)
at System.Net.TlsStream.CallProcessAuthentication(Object state)
at System.Threading.ExecutionContext.RunInternal(ExecutionContext executionContext, ContextCallback callback, Object state, Boolean preserveSyncCtx)
at System.Threading.ExecutionContext.Run(ExecutionContext executionContext, ContextCallback callback, Object state, Boolean preserveSyncCtx)
at System.Threading.ExecutionContext.Run(ExecutionContext executionContext, ContextCallback callback, Object state)
at System.Net.TlsStream.ProcessAuthentication(LazyAsyncResult result)
at System.Net.TlsStream.Write(Byte[] buffer, Int32 offset, Int32 size)
at System.Net.PooledStream.Write(Byte[] buffer, Int32 offset, Int32 size)
at System.Net.ConnectStream.WriteHeaders(Boolean async)
— End of inner exception stack trace —
at System.Net.HttpWebRequest.GetResponse()
at NzbDrone.Common.Http.Dispatchers.ManagedHttpDispatcher.GetResponse(HttpRequest request, CookieContainer cookies) in C:\BuildAgent\work\5d7581516c0ee5b3\src\NzbDrone.Common\Http\Dispatchers\ManagedHttpDispatcher.cs:line 117
at NzbDrone.Common.Http.Dispatchers.FallbackHttpDispatcher.GetResponse(HttpRequest request, CookieContainer cookies) in C:\BuildAgent\work\5d7581516c0ee5b3\src\NzbDrone.Common\Http\Dispatchers\FallbackHttpDispatcher.cs:line 53
at NzbDrone.Common.Http.HttpClient.ExecuteRequest(HttpRequest request, CookieContainer cookieContainer) in C:\BuildAgent\work\5d7581516c0ee5b3\src\NzbDrone.Common\Http\HttpClient.cs:line 121
at NzbDrone.Common.Http.HttpClient.Execute(HttpRequest request) in C:\BuildAgent\work\5d7581516c0ee5b3\src\NzbDrone.Common\Http\HttpClient.cs:line 76
at NzbDrone.Common.Http.HttpClient.Get(HttpRequest request) in C:\BuildAgent\work\5d7581516c0ee5b3\src\NzbDrone.Common\Http\HttpClient.cs:line 264
at NzbDrone.Core.Indexers.Newznab.NewznabCapabilitiesProvider.FetchCapabilities(NewznabSettings indexerSettings) in C:\BuildAgent\work\5d7581516c0ee5b3\src\NzbDrone.Core\Indexers\Newznab\NewznabCapabilitiesProvider.cs:line 57

I’ve found that restarting Sonarr will solve the issue temporarily, but I typically have to do this workaround at least every day unfortunately.

Yes. I setup a job to do restart sonarr. And it seemed to help for a couple days. Then stopped working. Then I setup a job to just restart the whole server. That also seemed to work for a couple days. Now, neither work and it just won’t connect to any indexers. So weird.

1 Like

From which browser? IE11 will be the closest to Sonarr’s behaviour.

The only other thing would be the lack of .net 4.5 leading to SSL issues, but I think that issue was different.

Just adding another voice of someone who’s having to do this. After a restart (and then “testing” the indexers) it works for a few days and then craps out again with the log message “The request was aborted: Could not create SSL/TLS secure channel.”

I ended up redeploying the entire thing (from OS to all apps). I backed every up. Reinstalled. Restored the backup. Now things are fine. So far.

I tried that too. No luck, still having the same problems.

Yea, after a Windows patch, it appears I am now having the issue again. And since the Sonarr folks aren’t really responding here, I’m not sure which update did it or how it affects this app.

I won’t be reinstalling or donating to Sonarr going forward. Just can’t get it to work any longer on Window and I’m not going to reinstall everything again without knowing what is causing the issue. I’ve asked here and on Discord just to be ignored. The only response was a vague reference in this thread to .net 4.5 without any real information on what is causing it.

I removed update KB4103723 and now it is connecting to indexers. Your mileage my vary. But, I’ll be moving away from Sonarr at this point.

Another plus 1 to this issue. I thought it originally had to do with Net issues, but that is not the problem. It reports none of the many indexers I have are connecting, yet testing each one, one by one, works fine. If I walk the list of indexers they stay OK for maybe a day or so. And then it is walk the isy again. Restarting Sonarr does not eliminate the problem.

Some on developers, throw us a bone on this one please. Overall, I am happy with the way it works, but this indexer things is annoying as hell.

To what, if I may ask?

When I first started this I was using Sickbeard for a long time with success. I moved to Sonarr due to folks on Reddit liking it. So, I’ll probably move to Sickrage (fork of Sickbeard). Sadly, I was always willing to pay for Sonarr. But, a crippling bug is causing this and with no real engagement from the Sonarr folks really shows where my money will go going forward. Easy decision. It’s not like I am the only one having the issue.

1 Like

Thanks. I will check that out. I agree, this was always a good platform, but it does not seem to be being supported right now. Hopefully we are wrong! :slight_smile:

Damn, installing Sickrage on Windows 7 is a bear. Simply refuses to work. Updated Python, installed PIP updates, tried installing Sickrage, and it just keep looping back to “Select your Python install folder” which works the first time, but when it loops back, it refuses to recognize the folder anymore. SIGH!

Simply put this is not something in Sonarr that is failing.

Similar issues were due to a lack of .net 4.5 being installed, which is about as specific as I can get, because you didn’t mention whether or not it was installed nor my direct question about which browser you used.

The fact that it started working after a full install leads me to believe something was missing, but then it came back (and nothing in Sonarr had changed, because v2 hasn’t been updated in months). Later removing KB4103723 from 2018 doesn’t make a lot of sense to me, that’s been installed countless times and doesn’t have any obvious links to .net and some seemingly unrelated ones to IE/Edge.

If you’re weighing in with “me too” please provide version information, OS, relevant logs, otherwise it doesn’t help troubleshoot.

Regarding donations, they’re for what is there, we don’t take donations for features or support.

If anyone reading this is interested in troubleshooting this issue, to start we’ll need to know:

  • All the requested information when you open a new thread (version, OS etc)
  • Enable trace logging
  • Which .net versions are installed
  • If you’re using https to connect to anything other than the indexers in Sonarr (download clients, etc)
  • How long from restart to failure typically
  • Check the logs for Failed to set TLS security protocol. or Disabled Tls1.2 due to remote certificate error.

If you want to discuss other applications you’ll need to do so elsewhere.

Maybe you could have responded to this… oh… I don’t know… OVER a month ago?

All you did in your first post was say something about .net 4.5 and how it was unrelated. Then you come back OVER a month later with different .net 4.5 information.

The only thing that sparked a response this time was the talk about a different solution.

Frustrating.

I should have said error message, instead of issue, but regardless the question was valid.

No it’s very much the same. I’m not going to chase people for information, if I’m asking questions and getting nothing for 10 days then the only response is about how it’s fixed it’s not something I’m keeping on top of, simple as that.

Good luck.

So I’m my fault you can’t communicate properly. Roger that.

You got nothing for 10 days because your first post was not a question. It was a statement. It wasn’t a post that was in any way helpful. You mention .net 4.5 but say it is unrelated. So what do you think people are going to think. I’m not the only one here that thought that or the others that posted would have done something with .net 4.5 also. Gimme a break with your posts.

This was the question, it was never answered.
10 days later everything was working, at that point things looked resolved and I stopped looking at replies, it wasn’t until there was a bunch of replies over a few days I saw it was still an issue.

We can only do so much dev/support, if that’s not good enough, then it’s not good enough, but I’m not going to argue about it, what’s done is done, if you want to troubleshoot it we can, if you’ve moved on, then we’ll drop it.

2 Likes

Not sure who you were answering here Markus. I am aware of the .NET issue, it is installed (4.8) and verified. What I see, most of the time, is all my indexers showing off line. But if I click on Test, they connect just fine, and if I do this to all of them, they all, (usually) say OK, and after all are done, the system alert badge goes away. It looks like the system, once it fails, does not recheck, or if it does, not often enough.

System info:

  • Windows 7 Ultimate, SP 1 64-bit OS
  • 8.00 Gb Installed memory, TB of disk available.
  • Microsoft .NET 4.8
  • https is used in a lot of places, including PLEX and Sannzbd, Sabnzbd server connections, and many websites. An SSH server also runs on the same machine.
  • Typical time from manual repair (restart does not always fix it) to failure about 24 hours.
  • No TLS failures, but I do see DNS failures. Except when I test DNS manually, it works fine! Example:
    20-10-19 13:14:02.6|Warn|Torznab|NZBGeek h20-10-1913:14:02.6|Warn|Torznab|NZBGeek https://api.nzbgeek.info/api?t=tvsearch&cat=5030,5040&extended=1&apikey=(removed)&offset=0&limit=100 DNS Name Resolution Failure: ‘api.nzbgeek.info’ DNS Name Resolution Failure: ‘api.nzbgeek.info’

When failures occurred, I have no issues navigating the web in a browser, and if I test, it works fine. Makes me wonder why DNS fails inside Sonarr?

Any of this helpful?

Thanks.

EDIT:
Oh, I tried to test today, and I got this and multiple indexers:
LOG
Clicking on test failed.

Restarting Sonarr did make this problem go away. Well. Sort of! After restarting, the badge and failed Indexers were still showing. But clicking on each of the listed failures connected properly and after all, the badge cleared.

One point of interest, I am on an always on VPN!

(Thanks for the merge of the two comments. Was just going back to do that when it disappeared! :slight_smile: )