I've been using netatalk for several years and it's always "just worked"
for me with little attention. However one of my servers began refusing to
grant write access to one of its shares with complaints in the log:
afpd: savevoloptions: cannot get lock: No locks available
cnid_metad: allocvolinfo("/volume_path"): No locks available
afpd: read: Connection reset by peer
On the connecting client, I would get a popup window with:
Something wrong with the volume's CNID DB, using temporary CNID DB
instead. Check server messages for details. Switching to read-only mode.
I tried combing google for information about this and found some
suggestions to remove various . directories created in the volume:
.AppleDB, .AppleDouble - which I did, but this did not help. The only
thing different about this volume is that awhile ago I set the group
sticky bit on the directory as it is also shared through samba and a
restricted group of users are allowed write access to the volume.
I've tried to remove all the apple . files/directories and thought
everything would just recreate itself when netatalk is restarted, but the
apple dot directories don't appear. I also tried to run "dbd -r" on the
volume: dbd -r /volume_path
but it reports: Not a Netatalk volume at '/volume_path', no .volinfo file
at '/volume_path/.AppleDesktop/.volinfo' or unknown volume options
It is a NFS mounted filesystem.
What creates the initial cnid database when netatalk is first turned on ?
Thanks for any help,