Re: bind timeouts

"Crist J. Clark" <cristjc_at_comcast.net> wrote:
> On Tue, May 11, 2004 at 11:36:35AM +0200, Cyrille Lefevre wrote:
> [snip]
> 
> The only obvious problem I see is,
> 
> [snip]
> 
> > * /etc/namedb/named.soa
> > 
> > _at_ IN SOA ns.gits.fr.invalid. root.gits.dyndns.org.  (


changed to gits.gits.fr.invalid

> > 2004051022 ; Serial
> > 3H ; Refresh
> > 1H ; Retry
> > 1W ; Expire
> > 1D ) ; Minimum
> > IN NS ns.gits.fr.invalid.

changed to gits.gits.fr.invalid

> [snip]
> > gits IN A 192.168.144.96
> [snip]
> > ns IN CNAME gits
> 
> A CNAME in an NS record is not a good idea, and I believe it is
> technically incorrect. Having one can cause weirdness.

nice try, but this doesn't fix anything :(

# time host -r localhost 127.0.0.1
Using domain server 127.0.0.1:
Host not found, try again.
   60.23s real     0.00s user     0.05s system
# time host -r localhost 127.0.0.1
Using domain server 127.0.0.1:
localhost.gits.fr.invalid has address 127.0.0.1
   10.07s real     0.01s user     0.02s system
# time host -r localhost 127.0.0.1
Using domain server 127.0.0.1:
localhost.gits.fr.invalid has address 127.0.0.1
    5.10s real     0.00s user     0.06s system
# time host -r localhost 127.0.0.1
Using domain server 127.0.0.1:
localhost.gits.fr.invalid has address 127.0.0.1
    0.06s real     0.00s user     0.04s system
# time host -r localhost 127.0.0.1
Using domain server 127.0.0.1:
(nothing!)
   20.11s real     0.01s user     0.04s system

the above times are the most common observed.

thanks anyway.

Cyrille Lefevre.
-- 
home: mailto:cyrille.lefevre_at_laposte.net