[2.5] Rack DNS config Failed to update DNS configuration and showed an UNHANDLED error instead of a nice error message why

[Andres Rodriguez]

I have a 4 node MAAS configuration with:

2 Region/racks (maas00, node01)
2 rack controllers (node03, node04)

I upgraded all of the machines, however, maas00, node03, node04 upgraded first, and successfully connected to the main region/rack (maas00), which is what /etc/maas/rackd.conf points to in node03/node04.

After the upgrade, I noticed the following traceback in the rack controllers (node03, node04):

2018-07-16 19:06:53 provisioningserver.rackdservices.dns: [critical] Failed to update DNS configuration.
        Traceback (most recent call last):
        Failure: twisted.protocols.amp.UnhandledCommand: (b'UNHANDLED', "Unhandled Command: b'GetDNSConfiguration'")

Once the node01 finished installing , i continued to see this issue, even in the logs of the region/rack rack.log. To fix the issue I had to:

1. In maas00,node01 I stopped maas-regiond/maas-rackd. Once stoped, I started the services again.
2. In maas00,node01 I started services again
3. restarted maas-rackd on node02,node03
4. No longer see the error messages.

The expectation is that errors like this should have correct error handling and MAAS shouldn't have had an unhandled error. Instead MAAS should have checked for what it expected, and since it didn't find it, show an error that implies the version mistmatch

[Blake Rouse (blake-rouse)]

This would be an upgrade error. The issue is that the rack controller is
using a newer version of code than the region controller.

That is not a bad error the error is actually correct. The region
controllers should always be upgraded before the rack controllers.

On Thu, Aug 9, 2018, 6:02 PM Andres Rodriguez <email address hidden>
wrote:

> ** Changed in: maas
> Assignee: (unassigned) => Blake Rouse (blake-rouse)
>
> --
> You received this bug notification because you are a bug assignee.
> https://bugs.launchpad.net/bugs/1782009
>
> Title:
> [2.5] Rack DNS config Failed to update DNS configuration.
>
> Status in MAAS:
> New
>
> Bug description:
> I have a 4 node MAAS configuration with:
>
> 2 Region/racks (maas00, node01)
> 2 rack controllers (node03, node04)
>
> I upgraded all of the machines, however, maas00, node03, node04
> upgraded first, and successfully connected to the main region/rack
> (maas00), which is what /etc/maas/rackd.conf points to in
> node03/node04.
>
> After the upgrade, I noticed the following traceback in the rack
> controllers (node03, node04):
>
> 2018-07-16 19:06:53 provisioningserver.rackdservices.dns: [critical]
> Failed to update DNS configuration.
> Traceback (most recent call last):
> Failure: twisted.protocols.amp.UnhandledCommand: (b'UNHANDLED',
> "Unhandled Command: b'GetDNSConfiguration'")
>
> Once the node01 finished installing , i continued to see this issue,
> even in the logs of the region/rack rack.log. To fix the issue I had
> to:
>
> 1. In maas00,node01 I stopped maas-regiond/maas-rackd. Once stoped, I
> started the services again.
> 2. In maas00,node01 I started services again
> 3. restarted maas-rackd on node02,node03
> 4. No longer see the error messages.
>
> To manage notifications about this bug go to:
> https://bugs.launchpad.net/maas/+bug/1782009/+subscriptions
>

[Andres Rodriguez (andreserl)]

The solution for this is to show a proper error implying/asserting that the versions mismatch and they need to upgrade.

[Adam Collard (adam-collard)]

This bug has not seen any activity in the last 6 months, so it is being automatically closed.

If you are still experiencing this issue, please feel free to re-open.

MAAS Team