[nsd-users] problems sending reload xfrdtomain: Broken pipe

Wouter Wijngaards wouter at nlnetlabs.nl
Mon May 13 07:08:05 UTC 2019


Hi Pierre-Philipp,

Setting database: "" in the configuration may help.

But regardless, I would like to reproduce the error.  If you could
(off-list, privately) send me the zonefile?

My guess is that the content of the zonefile somehow gives a problem, if
the below configuration causes the crash.

About debug info, NSD has the option to emit more debug information,
with -V 5.  If compiled with debug enabled, also -F -1 -L 2 can emit
even more debug information (per-packet and per-zone-update, so that is
a lot).

Best regards, Wouter

On 5/12/19 2:30 PM, pierre-philipp braun wrote:
> Yolo
> 
> I've compiled 4.1.27 on netbsd-current (8.99) without any issue, then configured it with a single forward zone and I always end-up with this error no matter what.
> 
> nsd[22124]: nsd starting (NSD 4.1.27)
> nsd[22124]: ...stale pid file from process 28114
> nsd[22124]: setup SSL certificates
> nsd[22673]: /nsd.db: not cleanly closed 0
> nsd[22673]: can not use /nsd.db, will create anew
> nsd[22673]: zone nethence.com read with success
> nsd[22673]: zone nethence.com written to db
> nsd[16988]: problems sending reload xfrdtomain: Broken pipe
> nsd[16988]: did not get start signal from main
> 
> Same goes with remote-control enabled
> Same goes with more advanced chroot pathes
> 
> I found the previous thread from 2014 on that matter but getting rid of /var/db/nsd/nsd.db (nor /var/run/nsd/nsd.pid) beforehand does not help.  My case seems different.
> 
> What is xfrdtomain expecting, is it creating a socket file?  And why is this called anyway, as the following MWE does NOT mandate any zone transfers to happen?  What could be causing the problem?  How to debug this?
> 
> server:
>         verbosity: 3
> 
> remote-control:
>         control-enable: no
> 
> zone:   
>         name: "nethence.com"
>         zonefile: "%s.db"
> 
> 
> I also looked at ktrace/kdump (netbsd's equivalent for strace) and I see some
> 
>  12105      1 nsd      RET   setsockopt -1 errno 55 No buffer space available
> 
> Other daemon live fine on this system so I doubt this is a netbsd bug (although that's current)
> 
> And thank you for the nice piece of software
> 
> _______________________________________________
> nsd-users mailing list
> nsd-users at NLnetLabs.nl
> https://open.nlnetlabs.nl/mailman/listinfo/nsd-users
> 

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: OpenPGP digital signature
URL: <http://lists.nlnetlabs.nl/pipermail/nsd-users/attachments/20190513/dadd510a/attachment.bin>


More information about the nsd-users mailing list