Ask Your Question
0

[Solved] Gateway Timeout error: Cannot access Horizon on OpenStack Newton.

asked 2016-12-20 20:28:08 -0600

AmirV gravatar image

updated 2016-12-27 12:33:14 -0600

Hi. I'm trying to install OpenStack Newton using the documentations here: http://docs.openstack.org/newton/install-guide-ubuntu/ (http://docs.openstack.org/newton/inst...)

I successfully installed these services: Identity, Image, Compute, and Network. Now I try to install Dashboard and I do it EXACTLY the same as the instructions at: http://docs.openstack.org/newton/install-guide-ubuntu/horizon-install.html (http://docs.openstack.org/newton/inst...)

However, I cannot open the Horizon (http://controller/horizon). It does not give any errors at first, it's just stuck in loading and after a while, I receive Gateway Timeout error.

Please see /etc/openstack-dashboard/local_settings.py in here: http://paste.openstack.org/show/592974/

Thank you.

edit retag flag offensive close merge delete

Comments

Probably Apache can't reach the WSGI components that make up Horizon. Check /var/log/apache2, and check if any Horizon wsgi processes are running.

Bernd Bausch gravatar imageBernd Bausch ( 2016-12-20 21:19:07 -0600 )edit

Thank you very much. I will post the logs here. But if it is the cause of error, how can I fix it?

AmirV gravatar imageAmirV ( 2016-12-20 21:41:59 -0600 )edit

If the processes are not running, find out why. If they are, find out what prevents Apache from reaching them. The log files may contain clues.

Bernd Bausch gravatar imageBernd Bausch ( 2016-12-20 21:44:33 -0600 )edit

It seems wsgi works fine. There is just one error: [Thu Dec 22 01:36:03.125705 2016] [wsgi:error] [pid 14317:tid 140137276241664] [client 10.0.0.10:57518] Truncated or oversized response headers received from daemon process 'horizon': /usr/share/openstack-dashboard/openstack_dashboard/wsgi/django.ws

AmirV gravatar imageAmirV ( 2016-12-21 16:07:12 -0600 )edit
add a comment

6 answers

Sort by ยป oldest newest most voted
4

answered 2016-12-27 12:32:44 -0600

AmirV gravatar image

updated 2016-12-27 12:33:56 -0600

I solved the problem as below.

1) open the dashboard configuration file:

#gedit /etc/apache2/conf-available/openstack-dashboard.conf

2) Add the following line to the configturation:

WSGIApplicationGroup %{GLOBAL}

3) Reload the Apache

#service apache2 reload

BOOM! enjoy the horizon.

edit flag offensive delete link more

Comments

I happened to have have the same problem and this solution worked. Thanks!!

Mohd gravatar imageMohd ( 2017-05-08 04:57:37 -0600 )edit

It helped me as well on one of the setups which I recently did. But what I am surprised of that Dashboard of my another setup is working smoothly without it. Both the setups are of Newton Version.

exergy gravatar imageexergy ( 2017-05-31 09:16:17 -0600 )edit
add a comment
0

answered 2019-05-25 07:31:21 -0600

VINAY gravatar image

If the problem still persists after default setting delete the duplicate cells which are mapped by using nova-manage

edit flag offensive delete link more
add a comment
0

answered 2019-08-27 21:51:17 -0600

neman gravatar image

Solved my problem, thank you!

Antun

edit flag offensive delete link more
add a comment
0

answered 2016-12-22 03:19:47 -0600

silently gravatar image

updated 2016-12-22 03:21:01 -0600 

[root@vm20 httpd]# pwd
/var/log/httpd

[root@vm20 httpd]# ls
access_log  error_log  ssl_access_log  ssl_error_log  ssl_request_log

look the content of files as above .

edit flag offensive delete link more
add a comment
0

answered 2016-12-21 02:42:26 -0600

visolve gravatar image

@AmirV Could you please try after changing the keystone version on /etc/openstack-dashboard/local_settings.py in Line OPENSTACK_KEYSTONE_URL = "http://%s:5000/v3.0" % OPENSTACK_HOST

edit flag offensive delete link more

Comments

Hi. Still not working.

AmirV gravatar imageAmirV ( 2016-12-21 16:07:29 -0600 )edit

May i know your hardware configuration?

visolve gravatar imagevisolve ( 2016-12-22 00:31:42 -0600 )edit

Sure. I have two VMs on VMware workstation on Windows 10. The first VM is controller with 8 CPU cores and 6 GB of RAM, and the second one is the compute node with 2 GB of RAM and 2 CPU cores.

AmirV gravatar imageAmirV ( 2016-12-24 18:38:18 -0600 )edit

@AmirV Try with this OPENSTACK_KEYSTONE_URL = "http://%s:5000/v3" % OPENSTACK_HOST in /etc/openstack-dashboard/local_settings.py .For me this works fine.

Pramod gravatar imagePramod ( 2017-02-17 13:29:28 -0600 )edit
add a comment

Get to know Ask OpenStack

Resources for moderators

Question Tools

1 follower

subscribe to rss feed

Stats

Asked: 2016-12-20 20:27:29 -0600

Seen: 7,867 times

Last updated: Aug 27 '19

Related questions

horizon errors [closed]

Horizon Dashboard Over HTTPS (IceHouse & Centos 6.5)

Horizon (Pike) is showing 500 error. [closed]

Can't re-login to horizon, bad session_id cookie [closed]

Unable to login in openstack dashboard(Havana)

'project_id' error after logging into Dashboard

Adding a drop down tab on Dashboard

Self Service Project Creation

Instance running but not appearing

Dashboard doesn't work (ServiceCatalogException at /admin/, Invalid service catalog service: compute)