Gentoo Websites Logo
Go to: Gentoo Home Documentation Forums Lists Bugs Planet Store Wiki Get Gentoo!

Bug 413649

Summary: net-fs/netatalk-2.2.1-r3 broken after upgrade
Product: Gentoo Linux Reporter: Jose Quinteiro <gentoo>
Component: New packagesAssignee: Network Filesystems <net-fs>
Severity: normal    
Priority: Normal    
Version: unspecified   
Hardware: All   
OS: Linux   
Package list:
Runtime testing required: ---

Description Jose Quinteiro 2012-04-26 17:45:40 UTC
The upgrade from -r2 -r3 removed the netatalk.conf. The ebuild informs you of this and says

    elog "The old configuration file /etc/netatalk/netatalk.conf is no longer"
    elog "installed, and will be ignored. The new configuration is supposed"
    elog "to be done through individual /etc/conf.d files, for everything that"
    elog "cannot be set already through their respective configuration files."

However, no configuration files are installed in /etc/conf.d or anywhere else.

Now cnid_metad crashes:
estupido ~ # /etc/init.d/cnid_metad start
 * Starting cnid_metad ...                                                                                                                                                      [ ok ]
estupido ~ # /etc/init.d/cnid_metad status
 * status: crashed

The log says:

Apr 26 10:39:46 estupido cnid_metad[303]: tsockfd_create: no suitable network config localhost:4700

Google sez that's because I need "CNID_METAD_RUN=yes". This used to be in my /etc/netatalkd.conf. I tried to figure out what the magic file I have to create in /etc/conf.d might be. The init.d script only says:

: ${CNID_CONFIG:=-l log_note}


Which files should I create in /etc/conf.d? Shouldn't they be installed for me?

Reproducible: Always
Comment 1 Jose Quinteiro 2012-04-26 18:48:13 UTC
I don't know which version I was running before. I assumed it was -r2, but I now think that was a bad assumption.
Comment 2 Jose Quinteiro 2012-04-26 18:58:19 UTC
The problem was than an old cnid_metad was still running. This was preventing the new cnid_metad from grabbing port 4700 on localhost. The new initd script didn't know how to kill the old cnid_metad.  I killed it manually and restarted afpd and now everything works.

I don't know if this is still a bug. It would be nice if the upgrade process stopped the old daemon before upgrading, or if the new initd script new how to kill old daemons too. At least and elog would be nice so other people don't pull out their hair like I did.