Ask Your Question
0

no data received from console (juno)

asked 2015-09-28 16:56:20 -0500

RedCricket gravatar image

updated 2015-09-29 12:00:59 -0500

Hi,

I have set up a Juno deployment of OpenStack. I can launch my cirros image:

[root@ost-mgmt-el7-001 ~]# nova list
+-----------+----------+--------+------------+-------------+------------------------+
| ID         | Name     | Status | Task State | Power State | Networks              
+-----------+----------+--------+------------+-------------+------------------------+
| xxxxxxx | test-007 | ACTIVE | -          | Running     | public-59=10.29.103.13
+-----------+----------+--------+------------+-------------+------------------------+
[root@ost-mgmt-el7-001 ~]# ssh 10.29.103.13 -l cirros
cirros@10.29.103.13's password: 
Permission denied, please try again.
cirros@10.29.103.13's password: 
$ ls -l
total 0
$ pwd
/home/cirros
$ exit
Connection to 10.29.103.13 closed.

But when I attempt to access the console like so ...

[root@ost-mgmt-el7-001 ~]# nova get-vnc-console test-007 novnc
+-------+---------------------------------------------------------------------------------------------+
| Type  | Url                                                                                         |
+-------+---------------------------------------------------------------------------------------------+
| novnc | https://ost-el7.exaple.com:6080/vnc_auto.html?token=xxxxx |
+-------+---------------------------------------------------------------------------------------------+
[root@ost-mgmt-el7-001 ~]# curl -v https://ost-el7.exaple.com:6080/vnc_auto.html?token=xxxxx

... I get "Empty reply from server" ...

* About to connect() to ost-el7.exaple.com port 6080 (#0)
*   Trying 10.29.103.39...
* Connected to ost-el7.exaple.com (10.29.103.39) port 6080 (#0)
* Initializing NSS with certpath: sql:/etc/pki/nssdb
*   CAfile: /etc/pki/tls/certs/ca-bundle.crt
  CApath: none
* SSL connection using TLS_RSA_WITH_AES_128_CBC_SHA
* Server certificate:
*   subject: CN=ost-el7.exaple.com,O="Example Systems, Inc.",L=San Jose,ST=CA,C=US
*   start date: Feb 18 21:54:13 2015 GMT
*   expire date: Feb 18 21:54:00 2017 GMT
*   common name: ost-el7.exaple.com
*   issuer: CN=HydrantID SSL ICA G2,O=HydrantID (Avalanche Cloud Corporation),C=US
> GET /vnc_auto.html?token=5612703c-06cb-497f-a205-bbebbebb6d9c HTTP/1.1
> User-Agent: curl/7.29.0
> Host: ost-el7.exaple.com:6080
> Accept: */*
> 
* Empty reply from server
* Connection #0 to host ost-el7.exaple.com left intact
curl: (52) Empty reply from server

Same thing happens if I log on to Horizon and attempt to access the console that way.

I am not sure how to really debug this. I do not see anything in the /var/log/httpd/*log files that is obviously wrong.

UPDATE: Here's my log file, /var/log/nova/nova-novncproxy.log :

# cat /var/log/nova/nova-novncproxy.log 
2015-09-28 16:46:20.922 15365 INFO nova.console.websocketproxy [-] WebSocket server settings:
2015-09-28 16:46:20.923 15365 INFO nova.console.websocketproxy [-]   - Listen on 0.0.0.0:6080
2015-09-28 16:46:20.923 15365 INFO nova.console.websocketproxy [-]   - Flash security policy server
2015-09-28 16:46:20.923 15365 INFO nova.console.websocketproxy [-]   - Web server. Web root: /usr/share/novnc
2015-09-28 16:46:20.923 15365 INFO nova.console.websocketproxy [-]   - SSL/TLS support
2015-09-28 16:46:20.924 15365 INFO nova.console.websocketproxy [-]   - Deny non-SSL/TLS connections
2015-09-28 16:46:20.925 15365 INFO nova.console.websocketproxy [-]   - proxying from 0.0.0.0:6080 to None:None
2015-09-28 16:59:33.722 15365 INFO nova.console.websocketproxy [-] Got SIGTERM, exiting
2015-09-28 16:59:33.793 15365 INFO nova.console.websocketproxy [-] In exit
2015-09-28 16:59:33.793 15365 DEBUG nova.console.websocketproxy [-] Closing socket listening at 0.0.0.0:6080 vmsg /usr/lib/python2.7/site-packages/websockify/websocket.py:828
2015-09-28 17:01:01.350 819 INFO nova.console.websocketproxy [-] WebSocket server ...
(more)
edit retag flag offensive close merge delete

Comments

look into /var/log/nova for debug. nova-proxy in particular

capsali gravatar imagecapsali ( 2015-09-29 06:44:35 -0500 )edit

1 answer

Sort by ยป oldest newest most voted
1

answered 2015-09-29 21:34:02 -0500

Moe gravatar image

Hi, novnc proxy works very great,

every instance on compute nodes on the start initiates a VNC port starting from 5900 and incrementally goes up novnc proxy configured in compute nodes to communicate to the controller node (no-vnc-proxy service) this service is a sort of abstraction between hypervisors vnc services and the clients who has accessed the instances using Horizon and delivers the service using the port 6080,

me myself have the same problem like you when I configured the servers both controller and compute node sides and restart the services

I mean every service related to openstack it did not work, I know it seems to be a little unprofessional but in my case I restarted the compute node and it worked I think there may be some backend services like libvirt which is involved with instance initiation .

I hope this helps

edit flag offensive delete link more

Comments

I found I had to set ssl_only to false in my nova.conf. I wish I knew what I am doing wrong. Could it be certificates?

RedCricket gravatar imageRedCricket ( 2015-09-29 21:54:34 -0500 )edit

Your Answer

Please start posting anonymously - your entry will be published after you log in or create a new account.

Add Answer

Get to know Ask OpenStack

Resources for moderators

Question Tools

1 follower

Stats

Asked: 2015-09-28 16:56:20 -0500

Seen: 765 times

Last updated: Sep 29 '15