On Apr 27 2009, Mohammed Ejaz wrote:
Can it be possible through the bind, www records should work as failover, I
mean during the primary record unavailable and then it should go for next
www only, Pls. note that I don't want let they work as round robin function.
1. Primary www record pointing
The Best - use carp (VRRP) protocol for it or nginx proxy server.
Or you can use dynamic update for zone:
ping -c 5 your.host || nsupdate ...
Mohammed Ejaz wrote:
> Hi all,
>
>
>
> Can it be possible through the bind, www records should work as
> failover, I mean during the primary record un
In article ,
wrote:
> This is not the DNS job to check at the web service availability.
> You could make an external script that is testing for the service availibil=
> ity
> and change the dns accordingly, like (...) :
>
> web1 active ?
> yes : was it active at last test ?
> y
, April 27, 2009 10:00 AM
To: me...@cyberia.net.sa; bind-users@lists.isc.org
Subject: RE: Failover
This is not the DNS job to check at the web service availability.
You could make an external script that is testing for the service
availibility
and change the dns accordingly, like
This is not the DNS job to check at the web service availability.
You could make an external script that is testing for the service availibility
and change the dns accordingly, like (...) :
web1 active ?
yes : was it active at last test ?
yes : do nothing
no : set www
I belive this is the job of a load balancer and not something that can
be done in dns.
DNS has no way of following the resulting answer ip or host, let alone
knowing what protocol to follow on and how to determine what ones
definition of "up" is.
This type if check could be done in a basi
6 matches
Mail list logo