Issue with using nzbhydra2 with Mylar

[phairplay]

Hi,
I noticed the following error on Mylar when searching my comics via nzbhydra2

[ERROR CODE: 900] org.springframework.validation.BeanPropertyBindingResult: 1 errors
Field error in object 'newznabParameters' on field 'o': rejected value [xml]; codes [typeMismatch.newznabParameters.o,typeMismatch.o,typeMismatch.org.nzbhydra.mapping.newznab.OutputType,typeMismatch]; arguments [org.springframework.context.support.DefaultMessageSourceResolvable: codes [newznabParameters.o,o]; arguments []; default message [o]]; default message [Failed to convert property value of type 'java.lang.String' to required type 'org.nzbhydra.mapping.newznab.OutputType' for property 'o'; nested exception is org.springframework.core.convert.ConversionFailedException: Failed to convert from type [java.lang.String] to type [org.nzbhydra.mapping.newznab.OutputType] for value 'xml'; nested exception is java.lang.IllegalArgumentException: No enum constant org.nzbhydra.mapping.newznab.OutputType.xml

I contacted the Mylar creator who said it’s nothing his end.

Could you please help resolve this?

[theotherp]

That was fixed two days ago. Are you sure you're running the latest version?

[phairplay]

Yes running version v1.0.14

[theotherp]

Can't really explain that. It should work... Wait for the next release.

[theotherp]

1.0.15 is out, try that one.

[theotherp]

Should be fixed now.

[phairplay]

hi,
im now seeing a search by mylar yet it doesn't download the result. i dont know if i should send this to mylar.

2018-01-11 18:23:14	ERROR	Unable to download nzb from link: https://127.0.0.1/nzbhydra/api?apikey=myapi&t=get&id=3 [http://127.0.0.1:5075/nzbhydra/getnzb/api/-6802808242181296776?apikey=myapi]
2018-01-11 18:23:14	INFO	download-retrieved headers: {'content-length': '70', 'x-xss-protection': '1; mode=block', 'x-content-type-options': 'nosniff', 'set-cookie': 'XSRF-TOKEN=fff4c5ca-6af0-4483-9360-f3464c66c283; Path=/nzbhydra', 'expires': '0', 'server': 'nginx/1.8.0', 'connection': 'keep-alive', 'pragma': 'no-cache', 'cache-control': 'no-cache, no-store, max-age=0, must-revalidate', 'date': 'Thu, 11 Jan 2018 18:23:14 GMT', 'content-type': 'application/xml;charset=UTF-8'}
2018-01-11 18:23:14	INFO	Download URL: https://127.0.0.1/nzbhydra/api?apikey=myapi&t=get&id=3 [VerifySSL:False]

[theotherp]

There's no port there

[theotherp]

And ffs, post your debug infos like it says in the issue template.

[phairplay]

apologise
here is the debug.
i use nginx to reverse proxy which works with all my other programs. all use http://
nzbhydra-debuginfos-2018-01-11-18-47.zip
12.0.0.1/nzbhydra, exepct for lazylibrarian which uses https://mydomain.com/nzbhydra

trying both on mylar cause the same error.

[theotherp]

Here's how I interpret Mylar's log:

Mylar accesses Hydra at https://127.0.0.1/nzbhydra/... to search but finds a download that links to http://127.0.0.1:5075/nzbhydra/.... Note the difference in scheme. I guess that could cause problems.

Try setting the external URL to https://127.0.0.1/nzbhydra/ (in Hydra's main config section).

[theotherp]

I'll probably disable the XSRF protection in the next version because it only causes problems and is rarely needed.

[phairplay]

added the url but still getting the same error

[fryfrog]

I'm gonna suggest using your reverse proxy harder. I have mine setup so that sub.domain.com points at my external ip via real DNS. But inside my network, my router points sub.domain.com to the server's internal ip. That way, I can just use that everywhere. This could fix your issue because now you can point everything at https://sub.domain.com/nzbhydra2 (or whatever) and everything will just "work". Magically. Hopefully.

[phairplay]

With lazylibrarian I have to use https://mydomain.com/nzbhydra which works fine. Trying this in Mylar comes up with the same errror as using http://127.0.0.1/nzbhydra (this is used by sonarr and radarr)

[fryfrog]

Does your mydomain.com work correctly internally? With routers and port forwards and stuff, it can be a little weird to get working. I assume it does, since LL is working. I've never added Hydra to either, let me try now and see what it does.

Edit: It turns out both Mylar and LL were using my old NZBHydra, so I've switched them to the new. I see a search query go through fine from LL to NZBHydra. The Mylar queries make it through too, but I don't have anything good to search for and nothing is found. :/

[phairplay]

'LL works fine searches and results are sent to nzbget.
Mylar again searches but I get the error you see in my third post.
When I look into NZBHydra search history and run it again it pull results.

[phairplay]

hi,
i updated to the latest version v1.0.16 and i’m still seeing the following error

2018-01-12 14:53:21	ERROR	Unable to download nzb from link: http://127.0.0.1/nzbhydra/api?apikey=MYAPI&t=get&id=3 [http://127.0.0.1:5075/nzbhydra/getnzb/api/4065993004411482480?apikey=MYAPI]
2018-01-12 14:53:21	INFO	download-retrieved headers: {'content-length': '70', 'x-xss-protection': '1; mode=block', 'x-content-type-options': 'nosniff', 'expires': '0', 'server': 'nginx/1.8.0', 'connection': 'keep-alive', 'pragma': 'no-cache', 'cache-control': 'no-cache, no-store, max-age=0, must-revalidate', 'date': 'Fri, 12 Jan 2018 14:53:21 GMT', 'content-type': 'application/xml;charset=UTF-8'}
2018-01-12 14:53:20	INFO	Download URL: http://127.0.0.1/nzbhydra/api?apikey=MYAPI&t=get&id=3 [VerifySSL:False]

here are the debug logs
nzbhydra-debuginfos-2018-01-12-14-53.zip

I’ve made connect with @evilhero to see if he could assist from a Mylar point of view.

[theotherp]

Would be nice if mylar logged what exactly the error is...

…

On Fri, Jan 12, 2018 at 3:58 PM, phairplay ***@***.***> wrote: hi, i updated to the latest version v1.0.16 and i’m still seeing the following error 2018-01-12 14:53:21 ERROR Unable to download nzb from link: http://127.0.0.1/nzbhydra/api?apikey=MYAPI&t=get&id=3 [ http://127.0.0.1:5075/nzbhydra/getnzb/api/4065993004411482480?apikey=MYAPI ] 2018-01-12 14:53:21 INFO download-retrieved headers: {'content-length': '70', 'x-xss-protection': '1; mode=block', 'x-content-type-options': 'nosniff', 'expires': '0', 'server': 'nginx/1.8.0', 'connection': 'keep-alive', 'pragma': 'no-cache', 'cache-control': 'no-cache, no-store, max-age=0, must-revalidate', 'date': 'Fri, 12 Jan 2018 14:53:21 GMT', 'content-type': 'application/xml;charset=UTF-8'} 2018-01-12 14:53:20 INFO Download URL: http://127.0.0.1/nzbhydra/api? apikey=MYAPI&t=get&id=3 [VerifySSL:False] here are the debug logs nzbhydra-debuginfos-2018-01-12-14-53.zip <https://github.com/theotherp/nzbhydra2/files/1626419/nzbhydra-debuginfos-2018-01-12-14-53.zip> — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#51 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ANtAecIAgOmBzfzXwczmUUkN6xPcBw4Kks5tJ3MVgaJpZM4RaADR> .

[evilhero]

It does at the search level - the problem is that it's getting some kind of response from the 127.0.0.1/nzbhydra and is returning the retrieved headers response. So there's no error per-se from the response (there might be an error code, it doesn't trap that), but the response being retrieved doesn't contain the nzb.

Has nzbhydra2 changed the returning nzb url so that it doesn't have a 'searchresultid' anymore ? (that's what Mylar looks for to distinguish hydra from newznab queries in the past - it looks like the nzb trying to be retrieved above is just using the id field now)

[theotherp]

@phairplay What you get in the browser when you call http://127.0.0.1:5075/nzbhydra/getnzb/api/4065993004411482480?apikey=MYAPI ?

@evilhero Why does Mylar need to distinguish between newznab and Hydra? No, there's not searchresultid parameter anymore. I was not aware there were any programs that handled Hydra differently (or needed to do so in the first place).

[theotherp]

I can see errors in the log that indicate that some tool is trying to make an API call with the parameter t=7030 which doesn't make sense. Mylar seems to make queries using that category. Could that be related?

[theotherp]

Ahhh... Sorry, I'm blind.

http://127.0.0.1/nzbhydra/api?apikey=MYAPI&t=get&id=3 is not a a valid NZB URL. The ID would need to look like the one in http://127.0.0.1:5075/nzbhydra/getnzb/api/4065993004411482480?apikey=MYAPI (in that case 4065993004411482480).

Why does Mylar use that ID?

[evilhero]

It doesn't need to distinguish normally - but at some point (can't recall exactly when), the returning of the searchresultid was only occuring with nzbhydra, so I had to make an exception for the return call to be handled properly. Probably why it was trying to use the id of 3 instead of the other id.

If it's no longer needed, I can remove the searchresultid check and it should be able to retrieve it normally provided @phairplay can figure out his reverse proxy stuff.

[theotherp]

Weird, it seems to work with all the other tools. If you need my support let me know.

[evilhero]

No worries - this is all because I thought it'd be a good idea to allow for dynamic API replacement within the existing API download calls because of some edge case.

So Mylar basically just stores the id and then dynamically generates the url from the other info. Turns out to be more of a PITA than anything else - so now's the excuse for me to just remove that stuff entirely. I'll have this pushed out within the hour for @phairplay to try...

[phairplay]

Hi guys,
Many thanks for taking the time to review this, I’ll report back ASAP

[phairplay]

Winner winner chicken dinner.....
Thank you so much guys, I really appreciate it.

David

[theotherp]

Great to hear.