Taiga Github Integration gives 500 Server Error after accepting request with GitHub

330 views
Skip to first unread message

am...@polkaroo.net

unread,
May 9, 2016, 5:47:02 PM5/9/16
to taigaio

Hello. I just set up Taiga on Ubuntu 14.04 per the instructions here (without SSL so far): http://taigaio.github.io/taiga-doc/dist/setup-production.html

Logging in as admin worked.

After creating a GitHub developer application client ID, etc., following the GitHub integration instructions, and rebooting (not sure what to restart so went with a reboot), I was able to click the "sign in with github" link and authorize the application with github. But then when I'm sent back to my site, I wind up with some grumpy Ooompa Loompas ("Our Oompa Loompas have not been able to get you credentials from GitHub") and this error in the console:
angular.js:10722 POST http://taiga.gcrc.carleton.ca/api/v1/auth 500 (Internal Server Error)

Lots of these in celery.stderr.log:
[2016-05-09 21:36:14,156: ERROR/MainProcess] consumer: Cannot connect to amqp://taiga:**@127.0.0.1:5672//: [Errno 111] Connection refused.
Trying again in 2.00 seconds...

And this in gunicorn.stderr.log:


Any assistance would be appreciated.



Alejandro Alonso

unread,
May 10, 2016, 12:21:06 AM5/10/16
to am...@polkaroo.net, taigaio
Hello,

It sounds like your celery/asynchronous_tasks/rabbitmq connection isn't configured correctly, try to disable all that stuff in your local.conf to see if the error disappears. If it works after that you should review your config for the rabbitmq connection.

Regards!,

--
Please help us keep the Taiga.io Community open and inclusive, follow our Code of Conduct:
https://github.com/taigaio/code-of-conduct/blob/master/CODE_OF_CONDUCT.md
---
You received this message because you are subscribed to the Google Groups "taigaio" group.
To unsubscribe from this group and stop receiving emails from it, send an email to taigaio+u...@googlegroups.com.
To post to this group, send email to tai...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/taigaio/0c04dbe4-3380-43b7-aff5-477f39d2e30f%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.



--

  
Alejandro Alonso Fernández
CIO & Co-founder
www.kaleidos.net/FC8EAC/

Amos Hayes

unread,
May 10, 2016, 12:42:46 PM5/10/16
to Alejandro Alonso, taigaio
Thanks Alejandro. I commented out the following in taiga-back/settings/local.py and things are working. Apart from “SECRET01” which is changed to something else, these lines are right out of the documentation. The double forward-slash on the broker url is the only thing that looks odd, although it should make no difference as urls go. Where else should I be looking to see what is broken? Also, is there a faster way to restart everything other than rebooting? I’m used to using the OS mechanism to start/stop… are circus and ngnix the only two services I need to worry about running a ’service restart’ on?

#BROKER_URL = 'amqp://taiga:SECRET01@localhost:5672//'
#CELERY_RESULT_BACKEND = 'redis://localhost:6379/0'
#CELERY_ENABLED = True

#EVENTS_PUSH_BACKEND = "taiga.events.backends.rabbitmq.EventsPushBackend"
#EVENTS_PUSH_BACKEND_OPTIONS = {"url": "amqp://taiga:SECRET01@localhost:5672/taiga"}

Amos Hayes

unread,
May 10, 2016, 1:02:27 PM5/10/16
to Alejandro Alonso, taigaio
OK. I have it working now. It looks like the BROKER_URL needs to use guest:guest as the username and password. Should I be worried about the security implications of this?

David Barragán

unread,
May 10, 2016, 1:18:11 PM5/10/16
to Amos Hayes, Alejandro Alonso, taigaio
If depends on the configuration of your server/s and your infrastructure.


For more options, visit https://groups.google.com/d/optout.



--

---

 
David Barragán Merino  
Engineer

www.kaleidos.net/FFF8E7
Reply all
Reply to author
Forward
0 new messages