[Pdns-users] AXFR chunk error: Server Failure

[Steve Rose]

I am running a simple pdns and pdns-recursor 4.1.3 with auth server running
on port 5300 and the recursor on port 53 as per scenario 1 at
https://doc.powerdns.com/authoritative/guides/recursion.html, with one
master and one slave node in the cluster (both CentOS 7.3)

My slave is not receiving any zone updates and I am seeing the error below
on my slave in /var/log/messages ;

Aug 20 14:19:57 slave01 pdns_server[30600]: Aug 20 14:19:57 1 slave domain
needs checking, 0 queued for AXFR

Aug 20 14:19:57 slave01 pdns_server[30600]: Aug 20 14:19:57 Received serial
number updates for 1 zone, had 0 timeouts

Aug 20 14:19:57 slave01 pdns_server[30600]: Aug 20 14:19:57 Domain '
test.domain.com' is stale, master serial 2018082044, our serial 0

Aug 20 14:19:57 slave01 pdns_server[30600]: Aug 20 14:19:57 Initiating
transfer of 'test.domain.com' from remote '192.168.0.5'

Aug 20 14:19:57 slave01 pdns_server[30600]: Aug 20 14:19:57 Starting AXFR
of 'test.domain.com' from remote 192.168.0.5

Aug 20 14:19:57 slave01 pdns_server[30600]: Aug 20 14:19:57 Unable to AXFR
zone 'test.domain.com' from remote '192.168.0.5' (resolver): AXFR chunk
error: Server Failure

I am running bind backend on my slave, and the config looks like this ;

# cat /etc/pdns/pdns.d/pdns.backend.conf

launch=bind

bind-config=/etc/pdns/bindbackend.conf


# cat /etc/pdns/bindbackend.conf

zone "test.domain.com" {

  type slave;

  file "/var/sky/pdns/zones/stg3.devops.id.bskyb.com.zone";



  masters {

    192.168.0.5;

  };



};

I have created an empty file at /var/pdns/zones/test.domain.com.zone

# ls -l /var/pdns/zones/test.domain.com.zone

-rw-r--r--. 1 pdns pdns 0 Aug 20 09:09 /var/pdns/zones/test.domain.com.zone

The master node runs sqlite as a backend so I can leverage the pdnsutil
tool for managing the zone on the master. I am now not sure whether I can
run different backends on the master and the slave, I expected it not to be
an issue.

I am also wondering whether the changes in 4.1 linked to above are
contributing to this issue, aside from a few changes the config has been
ported from version 4.0 of powerdns.

How would you tackle the above error from here?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.powerdns.com/pipermail/pdns-users/attachments/20180820/1b58aa7b/attachment.html>