[Gluster-devel] BUG: $CONFIGFILE undefined in /etc/init.d/glusterfs

[Brandon Simmons]

the do_start() function has the line:

    start_daemon -p $PIDFILE $DAEMON -f $CONFIGFILE

where CONFIGFILE is not defined in the init script.

For reasons inexplicable, this only causes an error when I try to edit the file

    /etc/glusterfs/glusterd.vol

#> sudo /etc/init.d/glusterd start
 * Starting glusterd service glusterd    [ OK ]
/usr/sbin/glusterd: option requires an argument -- 'f'
Try `glusterd --help' or `glusterd --usage' for more information.

...after which glusterd is not running.

Perhaps glusterd only pays attention to the '-f' flag if there is an
error in glusterd.vol, which I guess is a default config file of some
kind. Hell if I know, since none of this seems to be documented
anywhere in an up to date or coherent fashion.

IYI, here is the /etc/glusterfs/glusterd.vol file which seems to
result in glusterd crapping out:

#> cat glusterd.vol
volume management
    type mgmt/glusterd
    option working-directory /etc/glusterd
    option transport-type socket,rdma
    option transport.socket.keepalive-time 10
    option transport.socket.keepalive-interval 2
end-volume

# Added by me!
volume test-vol-posix
  type storage/posix
  option directory /mnt/brick1
end-volume

volume locks
  type features/locks
  option mandatory-locks on
  subvolumes test-vol-posix
end-volume

volume brick
  type performance/io-threads
  option thread-count 4
  subvolumes locks
end-volume

volume server
  type protocol/server
  option transport-type tcp
  option auth.addr.brick.allow *
  subvolumes brick
end-volume

The above additions were taken from this out of date document:
http://www.gluster.com/community/documentation/index.php/NFS_Like_Standalone_Storage_Server

Brandon