[00:13] <Denyel> Not sure if anybody is on on Sunday, but I'm running out of ideas. I have a maas node booting bare metal Huawei boxes for commissioning, but commissioning fails.  The box boots, fails some metadata calls, and lands on the login screen
[00:14] <Denyel> I managed to unsquash and squash the filesystem so I have root access to the node, but don't know where to start looking beyond the logs
[00:35] <Denyel> The only errors are failed calls to get the instance-id 'http://10.20.59.1/latest/meta-data/instance-id'
[03:33] <pmatulis> Denyel, what about logs on the MAAS side? /var/log/maas/{maas.log,rackd.log,regiond.log}
[03:39] <pmatulis> what power type did you choose?
[15:11] <Denyel> Thanks for responding. The power was IPMI and it seems to power the boxes okay. I've used both LAN_1_5 and LAN_2_0. The rackd.log shows some tftp requests: pxelinux.0 twice, ldlinux.cfg, pxelinux.cfg twice, kernel, and initrd. It also shows the status transition and power state change in the maas.log.
[15:11] <Denyel> Power state changes from on to off, but it never tried to turn off again.
[15:12] <Denyel> Is the only thing that runs the cloud-*** files?
[16:42] <Denyel> I pasted the logs http://pastebin.com/tbzYB5K7
[16:44] <Denyel> The only things I see are 'cc_final_message.py [WARNING]: Used fallback datasource' from the node and a single 'could not serialize access due to concurrent update' from the postgreSQL server on the maas server
[16:46] <Denyel> I created a commissioning script '#!/bin/bash "root:password" | /usr/sbin/chpasswd' so I have access to the node.
[16:52] <Denyel> cloud-init.log http://pastebin.com/wmY8qPd7  and cloud-init-output.log http://pastebin.com/WFQA9jYX
[16:53] <Denyel> It looks like it finishes and the server never knows.
[19:23] <Denyel> I think I figured it out but don't know how to fix it. I have MAAS running on a system with two IP addresses. The can only access one, but the node's resolv.conf shows the IP it cannot access. How do I change the IP address nodes use to communicate with the MAAS server
[19:43] <Denyel> dpkg-reconfigure maas-region-controller
[19:43] <Denyel>  
[19:44] <Denyel> With the correct IP if somebody is looking through the #maas logs in the fiture
=== mup_ is now known as mup
[22:39] <DaToz_> hello can i have maas installed on 2 NICs one for management and one for PXE?
[22:51] <Denyel> Yes.
[22:52] <Denyel> That's how I did it. Just make sure that you configure the IP when installing with an IP you can hit from the nodes.
[22:58] <Denyel> One more problem for me. Now that I am getting the system to commission, my deployment of CentOS is hanging at the 'boot" prompt from PCELINUX. Ant ideas?