[PLUG] Redundant and backup
Josh Orchard
josh at emediatedesigns.com
Fri Oct 10 15:15:03 UTC 2003
> On Fri, 2003-10-10 at 12:33, Josh Orchard wrote:
>> Hello all,
>>
>> I hope your days are going well.
>>
>> I'm running a web server and have all the good stuff going on it such as DNS, Web,
>> FTP, Mail, etc. I have a second machine else where that does secondary DNS.
>>
>> This all works fine but would like to know if there is a way to set up some logic
>> in
>> that I can have the first machine backup up to the second and if the first goes
>> down
>> that the second will become the active machine? I know that there is something in
>> Mail that will do reduntant delivery or something. Some DNS MX record setting?
>>
>> Is there some DNS setting that I can achieve this for HTTP, FTP, and other
>> internet
>> services? I want to know if there is something in DNS that if it can't find one
>> web
>> server that it will go to another. Or maybe if it can't find one IP it will go to
>> another to try to resolve services. I will be backing up all the data remotely
>> over
>> rsync and such and would be nice to set up reduntant fall over systems via DNS.
>>
>> Is this possible? Do I make sense?
>
> The mechanism for doing this is DNS SRV (service) records. SRV
> records are just like an MCX records except that they're
> protocol-independent. MX records are only used with SMTP (so far),
> while SRV records can be used with any protocol including HTTP.
> Unfortunately, Web browsers do not support SRV records even
> though we've been trying to get the designers to support them for
> about 5 years.
>
> Ed
>
>
So, besides setting up yet another system inbetween the first two I do not have a
choice it seems. The Linux-HA concepts appears to need a server in front to do like
a load balancing concept. I was hoping for something like the DNS SRV records
concepts. It does make the most sense.
Hmm, I'll have to think about another solution. The two computers I have are not on
the same network and are located in two different places.
Thanks,
Josh
More information about the PLUG
mailing list