[dns-operations] "best practices" for restaring internal DNS servers

[Rubens Kuhl]

I said loss of cache, not loss of server, caused by a cache flush on the name server software. 

Rubens


Em 09/09/2012, às 20:38, Mark Andrews escreveu:

> 
> In message <7699449B-57AB-499C-93C2-DCF4049E42E9 at nic.br>, Rubens Kuhl writes:
>> 
>> It depends. I can tell from experience that for a large outbound SMTP syste=
>> m(1M+ accounts), loss of DNS cache can hit it pretty hard. =
> 
> Except we are not talking about the loss of a cache (recursive
> nameserver) which requires failover to a different cache.
> 
> We are talking about the difference to starting with a empty cache
> or a pre-populated cache and whether there is a significant difference
> from the clients perspective and for how many seconds there is a
> difference.  There is definitely some additional latency for the
> first few queries but after how many queries do you fall within the
> normal latency profile?
> 
> -- 
> Mark Andrews, ISC
> 1 Seymour St., Dundas Valley, NSW 2117, Australia
> PHONE: +61 2 9871 4742                 INTERNET: marka at isc.org