dig trace and "dig: Too many lookups"
Mark Andrews
Mark_Andrews at isc.org
Thu Jan 17 00:32:53 UTC 2008
> Our ns1 was down and a customer using an outsourced monitoring service
> forwarded me a report from DiG 9.3.1 doing a +trace. The ns2 (as far as I
> saw) was working fine.
>
> I don't know the dig command that was used but I am guessing:
>
> dig site.example.com +trace @some.root-servers.net
>
> since the output started with the root-servers.
>
> It has the same lookup repeated near 90 times. It looked like this:
>
> example.com 170877 IN NS ns1.example.net.
> example.com 170877 IN NS ns2.example.net.
> ;; Received 124 bytes from 65.77.130.5#53(ns2.example.net) in 79 ms
>
> example.com 170876 IN NS ns2.example.net.
> example.com 170876 IN NS ns1.example.net.
> ;; Recedig: Too many lookups
> ived 124 bytes from 65.77.130.5#53(ns2.example.net) in 80 ms
>
>
> The above "Recedig: Too many lookups" was in the report forwarded to me.
> The "dig: Too many lookups" was in the middle of the "Received 124 bytes
> ..." message.
>
> It had that a few times:
>
> ;; Receivedig: too many lookups
> ...
> ;; Recedig: Too many lookups
> ...
> ;; Recedig: Too many lookups
>
>
> I know this doesn't provide much details and the problem was from sixteen
> hours ago so I can't see it myself.
>
>
> Anyone seen this before? Any ideas what it means?
>
> Thanks!
>
>
> Jeremy C. Reed
>
> p.s. Hope it is okay I replaced the domain names above.
fatal("too many lookups"); -> stderr
printf(";; Received %u bytes from %s(%s) in %d ms\n\n", -> stdout
stdout was not flushed before before the stderr was output.
As a result part of the dtdout stream was before the message
to stderr and part after.
"dig +trace" does not try to track which servers it has
talked too, unlike a full iterative resolver which does
track this information. As a result it can easily loop
forever so there is a lookup limit which stop the lookup
when it starts to loop.
Mark
--
Mark Andrews, ISC
1 Seymour St., Dundas Valley, NSW 2117, Australia
PHONE: +61 2 9871 4742 INTERNET: Mark_Andrews at isc.org
More information about the bind-users
mailing list