Resolver timeouts, EDNS and networking

Christian Robottom Reis kiko at async.com.br
Thu Sep 27 23:23:28 UTC 2007


On Thu, Sep 27, 2007 at 04:20:09PM -0700, Chris Buxton wrote:
> If it were me, I would start by disabling EDNS0. This is often the  
> cause of problems such as this. You can try setting the buffer size  
> down to 512 bytes, but if that doesn't solve it, turn EDNS off entirely.
> 
> If that solves it, then probably the forwarders are the problem. Ask  
> your ISP about this; they may be using some kind of security software  
> that is not able to handle EDNS0.

Can I reliably detect whether my forwarders handle EDNS0 or not? I have
tried manually querying them using host and dig (with +bufsize) and they
seem to reply without a problem, but I'm not sure if that is a valid
test.

> There will always be some timeouts. That's just the nature of the  
> Internet today. But these should not result from a large percentage  
> of your server's resolving attempts.

Thanks for reassuring me. I'm trying hard to get better service to my
users -- and intermittent or random failures are particularly
bothersome.
-- 
Christian Robottom Reis | http://async.com.br/~kiko/ | [+55 16] 3376 0125



More information about the bind-users mailing list