Merge lp://qastaging/~danilo/maas/bug-1688060-houdini-pod-nodes into lp://qastaging/~maas-committers/maas/trunk
Status: | Merged |
---|---|
Approved by: | Andres Rodriguez |
Approved revision: | no longer in the source branch. |
Merged at revision: | 6053 |
Proposed branch: | lp://qastaging/~danilo/maas/bug-1688060-houdini-pod-nodes |
Merge into: | lp://qastaging/~maas-committers/maas/trunk |
Diff against target: |
54 lines (+17/-2) 2 files modified
src/maasserver/models/bmc.py (+2/-2) src/maasserver/models/tests/test_bmc.py (+15/-0) |
To merge this branch: | bzr merge lp://qastaging/~danilo/maas/bug-1688060-houdini-pod-nodes |
Related bugs: |
Reviewer | Review Type | Date Requested | Status |
---|---|---|---|
Newell Jensen (community) | Approve | ||
Andres Rodriguez (community) | Approve | ||
Blake Rouse | Pending | ||
Review via email: mp+324166@code.qastaging.launchpad.net |
Commit message
When resyncing/
Description of the change
When syncing machines, we instead look at all Nodes in the database with matching interface MAC addresses, and not just all Machine nodes (since rack controllers are not part of those).
As a side-effect, this will also match region-
Test instructions:
1. Register a (eg. virsh) pod
2. Compose a few machines
3. Deploy one machine
4. ssh to that machine and install maas-rack-
5. Go to Nodes/Controllers page and follow "add rack controller" instructions to add the deployed node
It will disappear from the listing of machines in a pod, but at least refreshing a pod won't cause it to recommission (and mess up your MAAS).
lgtm! I think this would prevent any machine that was part of the pod and became something else to be recommissioned if refreshed.
The one thing in the future is that we will have to show it there with a different status. But that's a future thing.
Since Newell/Blake worked on RSD, I'd like a review from either of them before landing.