Page MenuHomePhabricator

No DNS entry for deployment-logstash1.eqiad.wmflabs ?
Closed, ResolvedPublic

Description

deployment-logstash1.eqiad.wmflabs (from deployment-prep) seems to be missing its DNS entry:

root@deployment-restbase01:/# ping deployment-logstash1.eqiad.wmflabs
ping: unknown host deployment-logstash1.eqiad.wmflabs

However, using its IP address shows the host is actually ping-able:

root@deployment-restbase01:/# ping 10.68.16.134
PING 10.68.16.134 (10.68.16.134) 56(84) bytes of data.
64 bytes from 10.68.16.134: icmp_seq=1 ttl=64 time=0.691 ms
64 bytes from 10.68.16.134: icmp_seq=2 ttl=64 time=2.12 ms
^C
--- 10.68.16.134 ping statistics ---
2 packets transmitted, 2 received, 0% packet loss, time 999ms
rtt min/avg/max/mdev = 0.691/1.406/2.122/0.716 ms

Event Timeline

mobrovac raised the priority of this task from to Needs Triage.
mobrovac updated the task description. (Show Details)
mobrovac subscribed.

We narrowed down the problem to be local to the ldap-backed dns, @coren is looking into it at the moment.

The issue may actually lie within dnsmasq after all; the instance is properly listed in the list it should be serving the name of, yet gives a SRVFAIL rather than the A record on query. Digging to see if we reached some limit in dnsmasq.

Andrew claimed this task.
Andrew subscribed.

No idea why this broke, but a reboot fixed it.

As a further note (so that we can recognize the bug if it recurs), dnsmasq /did/ have a valid lease for that IP and had the correct name associated with it, and the reboot required a "hard reboot" (per nova)