[Openstack-operators] Inventory contains out of date data
Ian.Banks at serverchoice.com
Thu Dec 8 12:32:03 UTC 2016
This is partly related to a previous submission "Mitaka to Newton networking issues" by a colleague of mine, Grant Morley.
We had to rebuild our ansible server due to issues experienced when reverting from the Newton to Mitaka versions, which left ansible in an unworkable state. Our osa (/opt/openstack-ansible ) and openstack_deploy (/etc/openstack_deploy) are stored on a local git repository, which we pushed to the new ansible server and successfully deployed our Mitaka Openstack environment. Since the initial build however we have pushed some changes from our openstack_deploy repository, and this seems to have had the effect of overwriting the inventory with the hostnames and IPs of our previous Newton build.
We are now unable to run any playbooks from ansible as it is trying to connect to the incorrect hosts/IPs. Can anyone please advise if there is a way to regenerate the inventory from the live containers currently running. All attempts have resulted in the incorrect old inventory being generated.
Also can anyone advise where dynamic_inventory.py retrieves data to generate the inventory?
Senior Systems Engineer
Unit H, J and K
Please note we have new telephone numbers.
Direct: 01438 532 307
Office: 01438 532 300
Support: 01438 342 310
ian.banks at serverchoice.com
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the OpenStack-operators