[Pdns-users] pdns_recursor dumping core

[a b]

About three weeks ago, I finally finished most of the work of packaging and integrating PowerDNS on Solaris 10, then deployed it on a test zone (Solaris lightweight virtual server) in order to "soak" it before fully going into production.

Now it is starting to conk out with the following message:
Nov 06 01:41:46 Calling daemonize, going to background[ Nov  6 01:41:46 Method "start" exited with status 0 ][ Nov 10 09:48:37 Stopping because process dumped core. ][ Nov 10 09:48:37 Executing stop method ("/opt/powerdns/sbin/rec_control quit") ]Fatal: Unable to connect to remote '/var/run/pdns_recursor.controlsocket': Destination address required[ Nov 10 09:48:38 Method "stop" exited with status 1 ][ Nov 10 09:48:38 Executing stop method ("/opt/powerdns/sbin/rec_control quit") ]Fatal: Unable to connect to remote '/var/run/pdns_recursor.controlsocket': Destination address required[ Nov 10 09:48:38 Method "stop" exited with status 1 ][ Nov 10 09:48:38 Executing stop method ("/opt/powerdns/sbin/rec_control quit") ]Fatal: Unable to connect to remote '/var/run/pdns_recursor.controlsocket': Destination address required[ Nov 10 09:48:38 Method "stop" exited with status 1 ][ Nov 10 13:56:37 Leaving maintenance because clear requested. ][ Nov 10 13:56:37 Enabled. ][ Nov 10 09:48:38 Method "stop" exited with status 1 ][ Nov 10 13:56:37 Leaving maintenance because clear requested. ][ Nov 10 13:56:37 Enabled. ][ Nov 10 13:56:37 Executing start method ("/opt/powerdns/sbin/pdns_recursor --daemon") ]Nov 10 13:56:37 PowerDNS recursor 3.3 (C) 2001-2010 PowerDNS.COM BV (Apr  6 2012, 20:37:51, gcc 3.4.3 (csl-sol210-3_4-branch+sol_rpath)) starting upNov 10 13:56:37 PowerDNS comes with ABSOLUTELY NO WARRANTY. This is free software, and you are welcome to redistribute it according to the terms of the GPL version 2.Nov 10 13:56:37 Operating in 64 bits modeNov 10 13:56:37 Reading random entropy from '/dev/urandom'Nov 10 13:56:37 Only allowing queries from: 127.0.0.0/8, ::1/128Nov 10 13:56:37 Will not send queries to: 127.0.0.0/8, ::1/128Nov 10 13:56:37 Inserting rfc 1918 private space zonesNov 10 13:56:37 Listening for UDP queries on 127.0.0.1:5300Nov 10 13:56:37 Listening for TCP queries on 127.0.0.1:5300Nov 10 13:56:37 Calling daemonize, going to background[ Nov 10 13:56:37 Method "start" exited with status 0 ]
So after four days, the recursor process dumped core, as can be seen from the logs above. I wanted to investigate further by analyzing the core file, but I cannot find it anywhere on the system! The limit for "coredumpsize" is "unlimited":
coredumpsize    unlimited
1. how can I ensure that the next time pdns_recursor crashes, I get a core file, and where would it dump it?2. what does the error message "Fatal: Unable to connect to remote '/var/run/pdns_recursor.controlsocket': Destination address required" mean?3. is the error message in 2. a Red herring caused by the crash, and if so, why did SMF fail to restart it again (until I manually cleared the fault with `svcadm clear svc:/site/powerdns:recursor`? 		 	   		   		 	   		  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.powerdns.com/pipermail/pdns-users/attachments/20131110/5355e8d0/attachment.html>