And the winner:
** Claus Assmann <ca@informatik.uni-kiel.de>
> Why do you have such a short timeout (300sec) for the MX record?
> vicinity.com. 300 MX 10 carrera.vicinity.com.
> Make it much longer and it should have fewer problems (one succeeded
> lookup may last one week). And it causes less traffic :-)
We have web servers that we want to all appear as www.vicinity.com. So,
we set up Round Robin DNS to distribute the load amongst these servers.
The TTL was set to 5 minutes to try and maintain a balance on these
servers. However, this implementation is GREATLY flawed. I changed the
minimum(default) TTL to 1 day and the TTL for the hosts in the Round Robin to 5
minutes. This seems to have solved our problem! A much more
efficient/effective use of DNS, as well.
Additional thanks goes to:
---------------------------
Fedor Gnuchev <qwe@ht.eimb.rssi.ru>
<Daniel.Blander@ACSacs.Com>
---- Fedor Gnuchev <qwe@ht.eimb.rssi.ru>
> Dear David,
>
> try talking to ALTER.Net guys - it is their fault and you are helpless
> untill they'll checkout their router. No sense poking in DNS when IP
> packets evidently loop - and I bet you'll save lots of other people
> expiriencing same mysterios troubles.
>
> > 7 HssiX-0.SR1.EWR1.Alter.Net (206.157.77.98) 13.696 ms 14.016 ms
> 13.82
> ms
> > 8 Fddi0-0.CR2.EWR1.Alter.Net (137.39.33.99) 14.258 ms 14.784 ms
> 14.818
> ms
> > 9 105.Hssi4-0.CR2.SCL1.Alter.Net (137.39.58.10) 169.641 ms 200.306
> ms *
> > 10 * Fddi1-0.GW1.SCL1.Alter.Net (137.39.19.33) 78.758 ms 83.5 ms
> <<<> 11 proximus-gw.ALTER.NET (137.39.133.110) 85.912 ms * 83.8 ms
> > 12 * proximus-gw.ALTER.NET (137.39.133.110) 83.774 ms *
> > 13 * proximus-gw.ALTER.NET (137.39.133.110) 84.332 ms *
> > 14 * proximus-gw.ALTER.NET (137.39.133.110) 104.217 ms *
> <<<> 15 proximus-gw.ALTER.NET (137.39.133.110) 87.099 ms * 86.655 ms
> >
>
> With best regards
>
> Fedor Gnuchev
> (hm, or Ted - in this English-typing world...)
> # NB: Advice - ignore all advice by this author.
> mailto:qwe@ht.eimb.rssi.ru
>
Actually, we aren't letting in udp packets, except for DNS (et.al.?), so
this results on our router due to these filters.
-- "Daniel J Blander - Sr. Systems Engineer for ACS"
> If Solaris 2.5 is at use, have you applied any of the 2.5 Recommended
> patches?? If so, you may have missed or gotten hit with a bug that exists.
> Get the latest version of the DNS/BIND patches and read the readmes!! There
> are some critical caveats - additional patches that *must* be loaded. Also
> the sendmail patch is up to rev -07......otherwise, do what I originally did
> which was to backout the sendmail patch and all was ok....
>
thx,
dp
_____
\ | /icinity Corp. * David Andrew Pope
_ \_/ (408) 732-5948 * pope@vicinity.com _
[____________________|____________________]
This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:11:08 CDT