Cluster is disconnected after fresh install

[Andreas Hasenack]

I did a fresh install of 1.7.0~beta4+bzr3143-0ubuntu1~trusty1

Admin was created in the command line and I logged in on the web interface.

The clusters tab shows that the main cluster is disconnected, and the logs are full of these:
==> /var/log/maas/maas-django.log <==
ERROR 2014-10-01 14:29:40,005 maasserver Unable to get RPC connection for cluster 'scapestack'
ERROR 2014-10-01 14:29:40,008 maasserver Unable to get RPC connection for cluster 'scapestack'
ERROR 2014-10-01 14:29:40,012 maasserver Unable to get RPC connection for cluster 'scapestack'
ERROR 2014-10-01 14:29:40,027 maasserver Unable to get RPC connection for cluster 'scapestack'

==> /var/log/maas/maas-django.log <==
INFO 2014-10-01 14:28:54,656 twisted RegionServer connection established (HOST:IPv4Address(TCP, '127.0.0.1', 41597) PEER:IPv4Address(TCP, '127.0.0.1', 58691))
ERROR 2014-10-01 14:28:59,459 maasserver Unable to get RPC connection for cluster 'maas'
ERROR 2014-10-01 14:28:59,463 maasserver Unable to get RPC connection for cluster 'maas'
ERROR 2014-10-01 14:28:59,467 maasserver Unable to get RPC connection for cluster 'maas'
ERROR 2014-10-01 14:28:59,481 maasserver Unable to get RPC connection for cluster 'maas'
ERROR 2014-10-01 14:29:09,517 maasserver Unable to get RPC connection for cluster 'scapestack'
ERROR 2014-10-01 14:29:09,520 maasserver Unable to get RPC connection for cluster 'scapestack'
ERROR 2014-10-01 14:29:09,525 maasserver Unable to get RPC connection for cluster 'scapestack'

(it changed to "scapestack" after I changed the dns zone name).

[Andreas Hasenack (ahasenack)]

After issuing this command, it started working:

sudo service apache2 restart && sudo service maas-cluster restart

[Andreas Hasenack (ahasenack)]

I had to reconfigure the cluster now, because my first changes didn't "take" because of the disconnected problem.

[Andres Rodriguez (andreserl)]

Confirmed with bzr3155

[Andres Rodriguez (andreserl)]

Although.... the Cluster connected after a few seconds.. it wasn't immediately.

[Andreas Hasenack (ahasenack)]

It took way more than a few seconds. So long in fact that I went through the cluster configuration, dhcp ranges, dns zone name, etc, and it was still disconnected at the end. And it didn't connect.

After the restart, then it was quick. Maybe that's what you meant.

[Gavin Panella (allenap)]

I suspect this has the same underlying fault as bug 1373031, so I'm going to dupe it. We can un-dupe if it turns out I'm wrong.

[Mike McCracken (mikemc)]

I'm still seeing this with 1.7.0+bzr3299, which should have the fix for bug #1373031.

this is a fresh install in a trusty VM, and the root user was created via createsuperuser on the command line before hitting the web ui

neither restarting the apache2 and maas-clusterd services , nor rebooting the whole VM works to get the cluster connected.

# here's the end of the maas.log:

Dec 4 22:33:09 maas-server maas.lease_upload_service: [INFO] LeaseUploadService starting.
Dec 4 22:33:12 maas-server maas.bootresources: [INFO] Started importing of boot images from 1 source(s).
Dec 4 22:33:12 maas-server maas.bootresources: [INFO] Importing images from source: http://maas.ubuntu.com/images/ephemeral-v2/releases/
Dec 4 22:33:12 maas-server maas.bootsources: [INFO] Updated boot sources cache.
Dec 4 22:33:13 maas-server maas.bootresources: [INFO] Finished importing of boot images from 1 source(s).
Dec 4 22:33:24 maas-server maas.power_monitor_service: [ERROR] Cannot monitor nodes' power status; region not available.
Dec 4 22:33:24 maas-server maas.boot_image_download_service: [ERROR] Can't initiate image download, no RPC connection to region.
Dec 4 22:33:24 maas-server maas.dhcp.probe: [ERROR] Can't initiate DHCP probe, no RPC connection to region.
Dec 4 22:33:24 maas-server maas.lease_upload_service: [ERROR] Failed to connect to region controller, cannot upload leases
Dec 4 22:37:24 maas-server maas.lease_upload_service: message repeated 4 times: [ [ERROR] Failed to connect to region controller, cannot upload leases]
Dec 4 22:38:24 maas-server maas.power_monitor_service: [ERROR] Cannot monitor nodes' power status; region not available.
Dec 4 22:38:24 maas-server maas.boot_image_download_service: [ERROR] Can't initiate image download, no RPC connection to region.
Dec 4 22:38:24 maas-server maas.lease_upload_service: [ERROR] Failed to connect to region controller, cannot upload leases
(END)

#here's the end of the maas-django.log:

INFO 2014-12-04 22:33:49,589 twisted /usr/lib/python2.7/dist-packages/piston/handler.py:27: exceptions.UserWarning: Handler already registered for model BootResource, you may experience inconsistent results.
INFO 2014-12-04 22:33:49,589 twisted /usr/lib/python2.7/dist-packages/piston/handler.py:27: exceptions.UserWarning: Handler already registered for model BootSourceSelection, you may experience inconsistent results.
INFO 2014-12-04 22:33:49,590 twisted /usr/lib/python2.7/dist-packages/piston/handler.py:27: exceptions.UserWarning: Handler already registered for model BootSource, you may experience inconsistent results.
ERROR 2014-12-04 22:33:49,708 maasserver Unable to get RPC connection for cluster 'Cluster master' (master)
ERROR 2014-12-04 22:33:49,708 maasserver Unable to get RPC connection for cluster 'Cluster master' (master)
ERROR 2014-12-04 22:33:49,709 maasserver Unable to get RPC connection for cluster 'Cluster master' (master)
ERROR 2014-12-04 22:33:49,712 maasserver Unable to get RPC connection for cluster 'Cluster master' (master)
INFO 2014-12-04 22:33:55,392 twisted /usr/lib/python2.7/dist-packages/piston/handler.py:27: exceptions.UserWarning: Handler already registered for model BootResource, yo...

[Alessandro Festa (afesta)]

had same issue and found in my case I changed pserv.yaml uncommenting the following lines:

# resource_root: /var/lib/maas/boot-resources/current/
# port: 69

once added back the comments I've restarted apache2 and maas-clusterd services and everything came back to normal status