[bind10-dev] server crashed after reload - Secondary Server only, caching enabled...

[Michal 'vorner' Vaner]

Hello

On Thu, Sep 06, 2012 at 04:52:46PM -0500, Jeremy C. Reed wrote:
> > 2012-09-03 16:03:18.875 FATAL [b10-auth.auth] AUTH_SERVER_FAILED server
> > failed: RRSIG is being added, but no RR of covered type found: as34288.net.
> > 7200 IN RRSIG SSHFP 7 2 7200 20120917042044 20120831011030 43405 as34288.net.
> > zWGqGQr5AxLdZdX7+Xx6lwGGyBmT/jOR6H11ZXQ2/Qogv5zh86KO2x0XxnhCDSSSkQkAbVDNTqgx+uWVtaz10X8/BJXiAzJUOLGAB/iLKvPPI5tI2PJuTYfONy7LytWBkrSMmihUJcBXdmGKdNJ2edql6Y7PkX+z4deKSB9imF0=
> 
> 
> I don't understand why this is logged twice.

Maybe the server tried restarting and loadinig it again, again failing to load?

> > 2012-09-03 16:03:19.884 ERROR [b10-xfrin.config]
> > CONFIG_SESSION_STOPPING_FAILED error sending stopping message: [Errno 32]
> > Broken pipe
> > 2012-09-03 16:03:19.887 ERROR [b10-zonemgr.config]
> > CONFIG_SESSION_STOPPING_FAILED error sending stopping message: [Errno 32]
> > Broken pipe
> 
> And the order of shutdowns should not allow msgq or cfgmgr to exit 
> before other components. Or if that is okay, then at least they 
> shouldn't ever need to use msgq or cfgmgr in this case.

I don't think msgq ever exits before the others stop. There's just no „clean
stop“ code in msgq, msgq needs to be killed in the next shutdown stage. So
unless the boss started to send SIGTERMs already, msgq didn't exit, it crashed
(and because it has no logging, the messages got lost). I suspect it sometimes
can't handle too many components disappearing at once (some kind of SIGPIPE or
race condition there).

With regards

-- 
XML is like violence. If it doesn't solve your problem, use more.

Michal 'vorner' Vaner
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 198 bytes
Desc: Digital signature
URL: <https://lists.isc.org/pipermail/bind10-dev/attachments/20120907/6a70bfe3/attachment.bin>