Flexget 2.13.14 unhandled errors with transmission 2.93

[specie77]

Expected behaviour:

establish a connection with Transmission 2.93

Actual behaviour:

Everytime I execute flexget to connect to transmission, an unhandled error occurs. Transmission host was just upgraded to 2.93, and now flexget won't connect. Included settings.json from transmission for reference.
settings.json.txt

Steps to reproduce:

• Step 1:
  flexget execute --tasks clean_up

Config:

config.yml.txt

Log:

flexget.log

Additional information:

• FlexGet version: 2.13.14
• Python version: 3.5.3
• Installation method: pip
• Using daemon (yes/no): error occurs with and without daemon
• OS and version: Windows Server 2012 R2 Standard
• Link to crash log:
  crash_report.2018.04.25.225359801753.log

[lamski]

Have you tried updating transmissionrpc?

pip install --upgrade transmissionrpc

[specie77]

From what I can tell, 2.93 is the latest version, no?

[lamski]

I think Transmission 2.94 is out.

[specie77]

Done, upgraded. Now, for some strange reason, it's returning an incorrect username/password error. However when I connect via the web client, it works perfectly. Something is strange here.

[lamski]

Try explicitly adding the IP of the machine hosting flexget to the rpc-whitelist.
From the fix transmission/transmission#468

This Pull Request mitigates this attack by requiring a host whitelist for requests that cannot be proven to be secure, but it can be disabled if a user does not want security. I think users might not be find it intuitive that servers on localhost or their intranet are exposed to the web, but hopefully they will read the warning.

[lamski]

Wait, looking at your flexget config file, I think you need to set the transmission port to the rpc-port which, in your case is port 80. The transmission web port usually defaults to port 9091, but in your settings.json, it's set to port 80.

[specie77]

You’re right, it was port 9091, I annonimized the config. I was confused, it is the rpc port. But it tells me my username and password are incorrect, they are not.

[lamski]

Is your transmission running on the same machine as flexget?
Is transmission the desktop version or the server version?

Perhaps you could troubleshoot it using transmission remote GUI?

I have flexget and transmission running on an ubuntu server and transmission remains at 2.92 so I'm not experiencing the issue.

[specie77]

Transmission is actually running on a cloud hosted seedbox as an app. Unfortunately I don’t have access to the platform. It used to work well, then I don’t know what happened.

[lamski]

Found a similar issue in the trans-GUI issues list: transmission-remote-gui/transgui#1127

The fix seems to be to use the numeric IP address in the remote host property.

[specie77]

Wait, in the transmission settings, or my flexget config?

[lamski]

The transmission host in your flexget config.

From the transmission documentation, found these two new properties introduced in v 2.93:

• rpc-host-whitelist: String (Comma-delimited list of domain names. Wildcards allowed using ''. Example: ".foo.org,example.com", Default: "", Always allowed: "localhost", "localhost.", all the IP addresses. Added in v2.93)
• rpc-host-whitelist-enabled: Boolean (default = true. Added in v2.93)

If you still could not connect, try setting rpc-host-whitelist-enabled to false.

You could also check this issue from the transmission issue tracker: transmission/transmission#476

[FlexGet-Bot]

This issue has been mentioned on Forum - FlexGet. There might be relevant details there:

https://discuss.flexget.com/t/download-showrss-task-didnt-produce-any-entries-this-is-likely-due-to-a-mis-configured-or-non-functional-input/5196/6