[2.5] dhcpd.conf domain-name-servers has incorrect ip after removing hardcoded dns on subnet

Bug #1787032 reported by Andres Rodriguez
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
MAAS
Fix Released
Critical
Blake Rouse

Bug Description

I have a MAAS that has 4 maas machines all in the same subnet:

2 region/racks
2 racks controllers

maas00 - 10.90.90.1
node01 - 10.90.90.3
node02 - 10.90.90.4
node03 - 10.90.90.5

I have enabled DHCP in the subnet, and selected maas00 as the primary for DHCP. I've overridden the Subnet configuration with 10.90.90.3 as the DNS.

However, when I remove the override of 10.90.90.3 as the DNS for subnet 10.90.90.0/24, the configuration doesn't get automatically updated and the wrong DNS is in the config

Related branches

Revision history for this message
Andres Rodriguez (andreserl) wrote :

This could be an issue that the trigger is not working, or that MAAS is not correctly updating the dhcpd config.

Changed in maas:
importance: Undecided → Critical
milestone: none → 2.5.0
status: New → Triaged
description: updated
tags: added: rack-proxy track
no longer affects: maas/2.4
Revision history for this message
Andres Rodriguez (andreserl) wrote :

I had to manually restart rackd to update the config

summary: - [2.5] dhcpd.conf domain-name-servers has incorrect ip
+ [2.5] dhcpd.conf domain-name-servers has incorrect ip after removing
+ hardcoded dns on subnet
Changed in maas:
assignee: nobody → Blake Rouse (blake-rouse)
Changed in maas:
milestone: 2.5.0 → 2.5.0beta1
Revision history for this message
Blake Rouse (blake-rouse) wrote :

Did a lot of testing of this to ensure that the behavior I am seeing is correct. Seems that this bug is actually related to the other bugs fixed with DHCPD. The trigger do exist for this change, setting and unsetting the domain-name-servers on the subnet does update the DHCP config.

I did see the original behavior where it didn't change, but after 30 seconds the timeout kicks in and it was updated. So the work to ensure this keeps work has fixed this problem. But more work could be done to ensure this is faster, instead of the 30 second timeout and retry.

Changed in maas:
status: Triaged → Fix Committed
Changed in maas:
status: Fix Committed → Fix Released
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.