Ask Your Question

Magnum vm instance can ping controller ip but can't resolve it by hostname to notify heat of status

asked 2016-09-14 14:08:16 -0600

mitakauser gravatar image

updated 2016-09-15 13:18:44 -0600

Hi ,

We're currently implementing a magnum bay deployment for testing purposes. So far we've been able to create a bay successfully but not without a few quirks. I know that magnum calls the heat api to create the whole bay (master, minion, health monitor,lbaas,etc). I was getting curious as to why after 1 hour the bay status remained in "Create in Progress" and only one instance was created (baymodel config with 1 master and 1 minion) so I dove into the system logs (fedora 23) via journalctl and one thing caught my attention. This line in particular:

Aug 26 01:16:01 fe-tdlnn6z37o-0-j4pmkkqrwsjx-kube-master-mb5kf6rv4vzp.novalocal wc-notify[1580]: okcurl -i -X POST -H 'X-Auth-Token: gAAAAABXv5ZTcss3qrjgtVFM-f5zxrrEEC-Mx3Qbx7O_1fJUeqkkn0lM5omb9U3UvM-NBHS0pKKFcRJp0Apa7Zxz_MGBDssRdLUSA-e13hoEhBFOf-r3ZrXZ_4AegdbziOYSm6C23cYZZW_HU-pu2Fs6vpe4taFP_HaKOBHGdVA9wiaoq7qC99E' -H 'Content-Type: application/json' -H 'Accept: application/json' **http://controller:8004/v1/9d7a25b6e7a74bce9fc09f06a3b0e258/stacks/fedora-23-bay-76m7za4mekfi-kube_masters-mjtdlnn6z37o-0-j4pmkkqrwsjx/d558128b-96e3-4ddb-a8ba-f464995fd118/resources/master_wait_handle/signal** --data-binary '{"status": "SUCCESS"}'

In the magnum heat template directory (controller node) I found a script that makes this call, I'm assuming that this is to notify heat when the master node (and minion) creation was successful so heat can update its status back to cli/horizon. So from the instance I tried to ping controller first by ip address which was successful and then by hostname "controller" with output "unknown host controller". Only after I put in the entry in the /etc/hosts file (vm instance) the aforementioned call to the controller was able to notify heat of the master creation (same case as with the minions). My question is there a way to automate the name resolving for controller hostname without having to ssh into each node instance and modifying /etc/hosts?. Unless of course I'm missing something else here in which case any pointers would be greatly appreciated. Please also let me know if you need more information I'll be happy to provide it.

I followed this guide for installing magnum


edit retag flag offensive close merge delete


I got the same error as you. Have you solved the problem? Please let me know. Thanks!

JackLin gravatar imageJackLin ( 2016-09-29 00:32:50 -0600 )edit

1 answer

Sort by ยป oldest newest most voted

answered 2016-09-15 20:41:10 -0600

check your /etc/resolv.conf and make sure you can pull the record from the host with dig:

dig @dns.server

if you can do this, your host should be able to resolve it.

edit flag offensive delete link more

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


Asked: 2016-09-14 14:08:16 -0600

Seen: 186 times

Last updated: Sep 15 '16