Issue Type: | Bug |
---|---|
Priority: | 4 - Normal |
Status: | Resolved |
Created at: | 2015-04-13T17:51:08.000Z |
Updated at: | 2015-04-13T18:17:53.000Z |
Created by: | Former user |
---|---|
Reported by: | Former user |
Assigned to: | Former user |
Fixed: A fix for this issue is checked into the tree and tested.
(Resolution Date: 2015-04-13T18:17:53.000Z)
2015-04-16 Mars Climate Orbiter (Release Date: 2015-04-16)
A user in #manta on freenode reports that setup_manta_zone.sh is not working on a multi-node environment. Specifically, the "script is deploying manta0 on one of the compute nodes and after that running vmadm lookup -j alias alias=${ZONE_ALIAS}", which "will always fail as VM is running on another node".
This script uses SAPI to deploy the Manta zone, and must not be passing a server_uuid. Prior to NET-270, this would have caused the new zone to land on the headnode, which the script obviously expects. After NET-270, it's likely not to, unless DAPI happens to pick the headnode.
This should be pretty easy to fix by fetching the headnode's UUID from sysinfo and then passing that with the SAPI request.
sdc-headnode commit 204e38b (branch master, by Dave Pacheco)
MANTA-2633 setup_manta_zone.sh broken in multi-node deployments