Red Hat Bugzilla – Bug 1310746
"Select Hosts" page of "Create Cluster" wizard reports wrong number of disks for node which wasn't initialized when the wizard was started
Last modified: 2016-07-26 05:25:06 EDT
Created attachment 1129367 [details]
screenshot of Select Hosts page with a node with 0 disks reported by mistake
Description of problem
When "Create Cluster" task is started before the "Initialize Node" task
completes on all nodes, "2. Select Hosts" page of "Create Cluster" wizard
reports 0 disks detected for this node, even though this node actually
contains a disk drive. This happens because this node wasn't initialized when
the "Create Cluster" operation was started.
Version-Release number of selected component
Steps to Reproduce
1. Install skyring on server and prepare few hosts for cluster setup
2. Login to web interface and on the main page accept one node.
* Make sure you use node which has at least one disk available for usm
* One node is enough to reproduce the issue, moreover it allows you
to retry with the same cluster again.
3. Immediatelly click on "Create Cluster" button (so that the "Initialize Node"
task is not completed yet)
4. Stop on the "2. Select Hosts" page of "Create Cluster" wizard and check
number of disks reported for the node which has been just accepted.
The page reports "0 Disks" for the new node even though the machine has one
disk available and it has been successfully initialized in the mean time.
There are multiple solutions and I'm not sure which one would make more
sense wrt other usm use cases:
* Don't let to start "Create Cluster" until all "Initialize Node" tasks are
* The page reports that "Initialize Node" task is still running for this node
and so that the information for the new node are not available yet. In addition
to that, it would be good to update the list of nodes when all tasks finis.
Going "Back" in the wizard doesn't help. One has to start "Create Cluster"
again to see actual state of the new host.
Interesting fact: the "Storage Profiles" modal window shows correct number
of disks every time.
Note that USM doesn't even report the IP address of the node.
Nodes which are in 'active' state only will be listed in the cluster creation screen and hence the issue is resolved
I haven't seen this for long time. Last time tested with: