In article <(E-Mail Removed) tij.nl>,
Martijn Lievaart <(E-Mail Removed)> wrote:
>On Mon, 09 Oct 2006 06:48:20 +0000, Mark Andrews wrote:
>
>> In article <452790e4$0$32418$(E-Mail Removed)>,
>> Wolfgang Hercker <(E-Mail Removed)> wrote:
>>>Some weeks ago my mail server was put onto another machine with a
>>>different IP. After a few days most of the eMail traffic was redirected
>>>successfully to the new mail server.
>>>
>>>However there are still some eMails which are directed to the old mail
>>>server (which is still running).
>>>
>>>Why ?
>
>Probably because some spammers cache MX records. Don't ask me why, I've
>seen it happen more than once. Are you sure that is legitimate mail on the
>old MX?
>
>>>
>>>Is there a way to re-inforce the DNS re-location? In other words: Is
>>>there a way to force a re-propagation of the Mail-Server DNS change
>>>information all
>>>over the world wide Mail server structure?
>>
>> Firstly it doesn't "propagation all over the world". The only
>> propogation is between the DNS servers for the zone. The rest of the
>> world queries these servers when they want a answer.
>
>It does propagate all over the world. As long as the information is cached
>by parties they will use old information. The new information can thus be
>seen to propagate over the world. God knows I've seen it happen way to
>many times (hint set your TTL to 0 some time before making changes).
Given the way the OP used the word propogate the only part
of the DNS protocol which remotely resembles progation is
a NOTIFY driven zone transfer.
Also given the time span involved (weeks) all caches should have
cleared any old cached data.
- As an optional step, check the TTLs of arriving data looking
for RRs with excessively long TTLs. If a RR has an
excessively long TTL, say greater than 1 week, either discard
the whole response, or limit all TTLs in the response to 1
week.
As far as I am aware, all the major caching servers do this.
>> This sounds like you also changed the nameservers for the zone but failed
>> to make the old servers serve the new content or to stop the old servers
>> serving the zone.
>
>Or, the new SOA for the zone is less than the old SOA. Or any of the more
>common DNS setup errors.
>
>M4
>--
>Redundancy is a great way to introduce more single points of failure.
|