Hello Winfried,
as far as we have seen, putting EDNS in the response is not mandatory. Auth was
changed recently to do it (http://wiki.powerdns.com/trac/changeset/2649) but
note the commit message is wrong - we saw no improvement in interaction with
BIND from it. The recursor supports EDNS0 buf
Hello Peter,
that explains a lot. Does the Recursor full support EDNS0? I ask because
there is no "OPT PSEUDOSECTION" in the dig output. Is this pesudosection
not mandatory?
Example:
dig +noanswer +nostats +edns=0 @nameserver
; <<>> DiG 9.8.1-P1 <<>> +noanswer +nostats +edns=0 @nameserver
;
Hello,
On Dec 5, 2012, at 14:16 , Peter van Dijk wrote:
> HOWEVER, if the packet cache is enabled, and the query that caused a packet
> cache entry happened to have EDNS, non-EDNS clients coming in after that will
> also get the big answer. I will write a ticket about this.
http://wiki.powerd
Hello Winfried,
On Dec 5, 2012, at 10:07 , abang wrote:
> I wonder how can a answer packet from our PowerDNS Recursor (3.4-pre) exceeds
> 512 bytes. I thought this is the limit and it should be truncated. The MSG
> SIZE in the example below is 701. Has someone a explanation for this?
dig, by
On 05.12.2012 10:07, abang wrote:
I wonder how can a answer packet from our PowerDNS Recursor (3.4-pre)
exceeds 512 bytes.
By means of DNS Extensions. http://tools.ietf.org/html/rfc2671
I thought this is the limit and it should be truncated.
The hard limit is 64kBytes, though most nameserv
Hi,
I wonder how can a answer packet from our PowerDNS Recursor (3.4-pre)
exceeds 512 bytes. I thought this is the limit and it should be
truncated. The MSG SIZE in the example below is 701. Has someone a
explanation for this?
Winfried
dig +notcp +ignore . NS @217.0.43.145
; <<>> DiG 9.8
