How to minimize the downtime in my case

[Kevin Oberman]

On Thu, Mar 14, 2013 at 10:29 PM, Manish Rane <manishr78 at gmail.com> wrote:

> So the TTL value we are discussing here are individual NS TTL Value? Or
> the SOA Default TTL Value.
> When I viewed my ISP record I found that the SOA Default TTL Value is 12
> days and NS RR TTL Value is 3600 secs
>

The SOA does not and never has contained a default TTL value. It once
contained a minimum TTL value and it is still generally referenced as such,
but it has nothing to do with the TTLs of the records in the zone. In your
zone files every RR has a distinct TTL. That is what we are talking about.

What? You say that your zone file does not have a TTL on every record?
Well, not any more. It is usually specified using the $TTL macro. That
defines a default TTL for a zone. You only need specify a TTL when you want
something other than the default.

If the SOA used to specify a minimum TTL, what is it used for today?  It is
the time a negative cache entry should be held. It has been tht way since
1998 as per RFC2308. (A negative cache entry is created by the return of a
No Data or Non-existent Domain response and the so called minimum TTL is
returned with these errors.  It should normally be rather short... A few
minutes.

The TTL on most stable RRs should be hours or even days. You shorten the
TTL when you plan some change in a "permanent" record.
-- 
R. Kevin Oberman, Network Engineer
E-mail: rkoberman at gmail.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.isc.org/pipermail/bind-users/attachments/20130316/f41c248d/attachment.html>