Semantic check slowing down the storage infrastructure deployment
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
StarlingX |
Fix Released
|
Medium
|
Ovidiu Poncea |
Bug Description
Brief Description
-----------------
In 2+2+2, we should allow unlocking worker even if storage nodes are not in 'available' state.
@Brent: Given the applications are decoupled from the infrastructure, we should remove this check. This unnecessarily slows down the infrastructure deployment.
Message displayed to user:
Error: Unable to unlock hosts: compute-1, compute-2, compute-3. Reason/Action: Can not unlock a worker node until at least one storage node is unlocked and enabled.
Severity
--------
Minor: Deployment workflow improvement
Steps to Reproduce
------------------
1. Install in standard mode till both controller are available. Do not configure OSDs.
2. Install worker nodes, do not configure worker node but skip ceph-mon configuration on worker node and do not install any storage nodes.
3. Unlock any worker node
Expected Behavior
------------------
If ceph-mon is configured on any worker node we allow unlocking workers
If there is any node of storage personality we allow unlocking worker nodes
=> Unlocking workers is denied only if there is no certainty about storage model used so that we force users into adding a ceph mon to the cluster.
Reproducibility
---------------
<Reproducible/
State if the issue is 100% reproducible or intermittent. If it is intermittent, state the frequency of occurrence
System Configuration
-------
Multi-node system, Dedicated storage
Changed in starlingx: | |
assignee: | nobody → Ovidiu Poncea (ovidiu.poncea) |
description: | updated |
Changed in starlingx: | |
status: | Triaged → In Progress |
tags: | added: stx.storage |
I believe this was already reported by Brent in https:/ /bugs.launchpad .net/starlingx/ +bug/1829765