Merge lp:~lamont/maas/bug-1567249-1.9 into lp:maas/1.9
Status: | Merged |
---|---|
Approved by: | LaMont Jones |
Approved revision: | no longer in the source branch. |
Merged at revision: | 4590 |
Proposed branch: | lp:~lamont/maas/bug-1567249-1.9 |
Merge into: | lp:maas/1.9 |
Diff against target: |
79 lines (+35/-1) 3 files modified
docs/changelog.rst (+2/-0) src/maasserver/clusterrpc/power_parameters.py (+15/-1) src/maasserver/clusterrpc/tests/test_power_parameters.py (+18/-0) |
To merge this branch: | bzr merge lp:~lamont/maas/bug-1567249-1.9 |
Related bugs: |
Reviewer | Review Type | Date Requested | Status |
---|---|---|---|
Blake Rouse (community) | Approve | ||
Review via email: mp+299608@code.launchpad.net |
Commit message
Complain intelligently if the rack controller is running an old version (1.7) of maas (missing_packages was introduced in revno 4343)
Description of the change
Complain intelligently if the rack controller is running an old version (1.7) of maas (missing_packages was introduced in revno 4343)
A 1.7 cluster controller would successfully connect to a 1.9 region controller, and the first that anyone would have any issues was when they went to look at a node (not cluster/rack controller..) and got a KeyError traceback. Detecting this error and giving the admin a hint of why they're seeing this strange thing keeps then from diving down the rabbit hole.
With a 2.0 region controller, the 1.7 (and 1.9) cluster/rack controllers fail to connect at all, so it becomes clear that "there's an issue with cluster controller $FOO" rather quickly, and the obvious investigative actions lead to the right problem being found.
Looks good, but no test?
The description also makes since and why this is not needed in 2.0+.