Dns-operations mailing list

[Jeff Taylor <shdwdrgn AT sourpuss.net>]

All right, here's the deal...  As stated before, the serial on the oz
zone was reset, and DNS services do not download the new file unless the
serial is higher.  From what I understand, the new file will be forced
to download after the zone expires, which in this case is two weeks.  At
this point, all of the T1 servers that are listing the oz zone are now
up to date.

If your T2 is slaving the oz zone (bind9 setup) you can delete the zone
file and restart bind.  This will immediately force your server to
download the correct copy of the zone file, and you will be back in sync
again.

If you are running one of the other dns services that gather answers
directly from a T1 server, it is likely that you have a large cache.  If
you flush the cache, it will force the request of updated data.

The issue that was causing failures of the dns.opennic.glue zone has
corrected itself, and I do not currently see any servers failing that
zone.  Once we get past the issue with oz, everything should be fine.


On 06/22/2012 02:51 PM, Jeff Taylor wrote:
> The serial on the oz zone was reset to a lower value last night, so it
> is taking some time for all of the servers to get the correct file.  The
> version with ns10.opennic.glue in the SOA is the correct file, and the
> errors should clear up as we get the T1 servers back in sync.  In the
> meantime, you can ignore that specific error.