Webvirtmgr Error after clean install

Rodrigo_Vieira_da_Co · December 9, 2017, 7:07pm

NethServer Version: 7.4 1708
Module: webvirtmgr

Hi!
I downloaded a new iso (7.4 1708) and installed webvirtmgr, but the message below appears:

I tried even with an old iso (7.3 1611) and gave the same error, do not load the webvirtmgr nor start the service.
I also noticed that the following command no longer finds the package groups:
yum groupinstall --setopt=centos-base.enablegroups=1 --setopt=centos-updates.enablegroups=1 install basic-desktop desktop-platform x11 fonts -y

What can it be?

mrmarkuz · December 9, 2017, 8:41pm

Hi @Rodrigo_Vieira_da_Co,

installed nethserver-webvirtmgr on a 7.4 VM I used for tests and on a fresh VM and had similar error:

[root@testserver ~]# systemctl status webvirtmgr
● webvirtmgr.service - WebVirtMgr Gunicorn Server
   Loaded: loaded (/usr/lib/systemd/system/webvirtmgr.service; enabled; vendor preset: disabled)
   Active: failed (Result: exit-code) since Sat 2017-12-09 21:17:47 CET; 54s ago
  Process: 629 ExecStart=/usr/bin/python /usr/lib/python2.7/site-packages/webvirtmgr/manage.py run_gunicorn -c /usr/lib/python2.7/site-packages/webvirtmgr/conf/gunicorn.conf.py (code=exited, status=1/FAILURE)
 Main PID: 629 (code=exited, status=1/FAILURE)

Dec 09 21:17:46 testserver.domain.local systemd[1]: Started WebVirtMgr Gunicorn Server.
Dec 09 21:17:46 testserver.domain.local systemd[1]: Starting WebVirtMgr Gunicorn Server...
Dec 09 21:17:47 testserver.domain.local python[629]: Traceback (most recent call last):
Dec 09 21:17:47 testserver.domain.local python[629]: File "/usr/lib/python2.7/site-packages/webvirtmgr/manage.py", line 8, in <module>
Dec 09 21:17:47 testserver.domain.local python[629]: from django.core.management import execute_from_command_line
Dec 09 21:17:47 testserver.domain.local python[629]: ImportError: No module named django.core.management
Dec 09 21:17:47 testserver.domain.local systemd[1]: webvirtmgr.service: main process exited, code=exited, status=1/FAILURE
Dec 09 21:17:47 testserver.domain.local systemd[1]: Unit webvirtmgr.service entered failed state.
Dec 09 21:17:47 testserver.domain.local systemd[1]: webvirtmgr.service failed.

/var/log/messages

Dec 10 08:19:52 testserver esmith::event[15810]: [INFO] webvirtmgr-console has been started
Dec 10 08:19:52 testserver esmith::event[15810]: Event: nethserver-webvirtmgr-save SUCCESS
Dec 10 08:19:52 testserver webvirtmgr-console: WARNING:root:No local_settings file found.
Dec 10 08:19:52 testserver webvirtmgr-console: Traceback (most recent call last):
Dec 10 08:19:52 testserver webvirtmgr-console: File "/usr/lib/python2.7/site-packages/webvirtmgr/console/webvirtmgr-console", line 22, in <module>
Dec 10 08:19:52 testserver webvirtmgr-console: from vrtManager.connection import CONN_SSH, CONN_SOCKET
Dec 10 08:19:52 testserver webvirtmgr-console: File "/usr/lib/python2.7/site-packages/webvirtmgr/vrtManager/connection.py", line 13, in <module>
Dec 10 08:19:52 testserver webvirtmgr-console: from django.conf import settings
Dec 10 08:19:52 testserver webvirtmgr-console: ImportError: No module named django.conf
Dec 10 08:19:52 testserver systemd: webvirtmgr-console.service: main process exited, code=exited, status=1/FAILURE
Dec 10 08:19:52 testserver systemd: Unit webvirtmgr-console.service entered failed state.
Dec 10 08:19:52 testserver systemd: webvirtmgr-console.service failed.

So I installed django:

yum install django

Now the services are starting but WebVirtMgr like

Played around with this but no success:

Latest upstream update dates of virt packages:

Dec 05 08:14:05 Updated: libvirt-daemon-3.2.0-14.el7_4.4.x86_64
Dec 09 13:04:06 Updated: libvirt-daemon-3.2.0-14.el7_4.5.x86_64

Tried downgrading but no luck.

Rodrigo_Vieira_da_Co · December 10, 2017, 1:46pm

Hi Markuz!
I also had installed django (python), and even with downgrade is presenting problems … I believe it is a bug in the repositories.
I hope a miracle this week, because I have a server to deliver urgently!
Hugs and we’ll wait for the solution!

mrmarkuz · December 12, 2017, 5:49pm

Maybe @giacomo or @edoardo_spadoni can help us here?

Christopher_Suttles · December 12, 2017, 8:59pm

I was able to fix this on a clean install.

First, install django if you haven’t already.

sudo yum install django

Now, if you have installed nethserver-webvirtmgr, remove it and all its direct dependencies.

sudo yum remove nethserver-webvirtmgr
sudo yum autoremove

Once that was done, I simple reinstalled webvirtmgr.

sudo yum install nethserver-webvirtmgr

That will pull back the dependencies, and after enabling webvirtmgr in the webui and creating a password, I was able to sign into webvirtmgr with no issue. Seems like django is missing as a dependency here and is needed for it to install correctly? Hopefully someone else can chime in.

mrmarkuz · December 12, 2017, 10:54pm

Thanks, yum autoremove worked, “signal-event nethserver-webvirtmgr-update” does also solve the “500 Internal Server error”.

The dependency for python-django is there:

[root@testserver ~]# rpm -qR webvirtmgr | grep django
python-django

But on install webvirtmgr grabs the python-django16 instead of python-django package.

Dependencies Resolved
================================================================================
 Package
================================================================================
python-django16

giacomo · December 13, 2017, 8:40am

I think that something changed on EPEL repository.

I will try to take a look on it later today.

giacomo · December 13, 2017, 5:57pm

I confirm that EPEL package has been changed, I tried to fix the spec but I couldn’t found a solution.

This is a quick and dirty fix after installation:

rpm -e --nodeps python-django16
yum install python-django -y
 rm -f /usr/lib/python2.7/site-packages/webvirtmgr/webvirtmgr.sqlite3
signal-event nethserver-webvirtmgr-update