Dns-operations mailing list

[Jeff Taylor <shdwdrgn AT sourpuss.net>]

The report pages are simply processing the information from each series
of server tests.  They list ALL servers because we may wish to know
which servers were being tested at any given date.  If we did not show
the "pass" status of each server, an operator may mistakenly assume
their server is working fine when in fact it was never tested.

The simplest method of keeping an eye on things is by monitoring
http://www.opennicproject.org/t2log/ and noting if your server(s) are
showing any errors.  If there is an error, you can either view the
details in the reports sent to the mailing list, or you can click on the
server name on the web page, and run an instant test on your machine to
determine where the failure is at.

If you see a failure for particular zones, troubleshooting should be
relatively easy - check that you are receiving the zone files if you
slave them, or confirm that the master server you are pointing at still
exists.  If you get any (recurs) errors, you are not allowing recursive
queries for that zone.

Errors that the server is unreachable are more difficult to
troubleshoot.  Obviously make sure your server is online and your dns
service is running.  Check that your firewall is allowing queries (both
udp and tcp!) through to your dns server.  If the t2log page is showing
a red bar for your server, then your machine cannot be reached from the
internet, so you need to verify your network configuration.  Beyond
that, I cannot give any troubleshooting tips because every server
configuration will be different.

As for your specific question about ns1.cz... There is no such server
currently operating.  I don't know when that server was removed, but it
is no longer a valid T2 server and should not be used.



On 06/10/2012 06:32 AM, opennic AT lewman.us wrote:
> On Thu, Jun 07, 2012 at 10:38:37PM -0500, alex AT opennicproject.org wrote 
> 0.5K bytes in 16 lines about:
> : It is pretty important that all T1/T2 operators are subscribed to both
> : so that is why this is happening.
>
> After reading the dnsreports, I have some thoughts:
>
> 1. Only send out the errors or failures. The default state should be
> reachable. The current state of the reports is training me to ignore
> them because 95% of it tells me "everything is working".
>
> 2. What is the action one is supposed to take if their server appears
> 'unreachable'? I ask because it seems ns1.cz/.micro haven't been reachable
> for days.
>