Page MenuHomePhabricator

cloudcontrols can't reach ns-recursor.openstack.eqiad1.wikimediacloud.org
Closed, InvalidPublic

Description

Currently the nova-fullstack test (running on cloudcontrol1006) validates the dns record of a new VM with dig @ns-recursor.openstack.eqiad1.wikimediacloud.org. That now results in 'connection timed out; no servers could be reached'.

That's maybe fine -- I can't really think of why external services should be able to talk to the recursor. I do note that cloudcontrol1006 can dig @ns-recursor0.openstack.eqiad1.wikimediacloud.org though.

Related Objects

StatusSubtypeAssignedTask
Resolvedaborrero
Resolvedaborrero
Resolvedaborrero
Resolvedaborrero
Resolvedayounsi
Resolvedcmooney
ResolvedPapaul
Resolvedcmooney
Resolvedcmooney
Resolvedaborrero
Resolvedaborrero
Resolvedaborrero
Resolvedtaavi
Opencmooney
Resolvedaborrero
Opencmooney
Resolvedaborrero
Resolvedaborrero
Resolvedaborrero
Resolvedaborrero
Resolvedaborrero
ResolvedAndrew
ResolvedAndrew
ResolvedAndrew
OpenAndrew
OpenAndrew
ResolvedAndrew
Resolvedaborrero
In Progresstaavi
OpenNone
OpenNone
Resolvedaborrero
Resolvedcmooney
Resolvedaborrero
Invalidaborrero

Event Timeline

Change 956497 had a related patch set uploaded (by Andrew Bogott; author: Andrew Bogott):

[operations/puppet@production] nova-fullstack: check dns via auth server rather than recursor

https://gerrit.wikimedia.org/r/956497

Change 956497 merged by Andrew Bogott:

[operations/puppet@production] nova-fullstack: check dns via auth server rather than recursor

https://gerrit.wikimedia.org/r/956497

cloudcontro1006/1007 should be re-racked and reimaged into the new network setup per the plans described at T341060: openstack eqiad1: introduce cloud-private and cloudlb.

All tests, checks, daemons and special scripts should be relocated into cloudcontrol1005 that, as of this writing, is the one in the new network setup (and can operate on cloud-private).