Monitor Locking Process

[kang at risca.rs.msu.edu]

> I am running GRASS4.1 on a linux machine.  To correct a suspected corruption
> error, I re-installed GRASS from the same binary files originally used.  I
> now recieve the message "could not complete locking process for monitor
> <x0>; lock file is /grass/locks/ansel/mon.106888" after attempting to run
> d.mon to start a monitor.
 
> The /grass/locks subdirectory exists, but only contains iwr_62/; no "ansel/"
> (ansel is the name of the linux box).  

For those who may have problem on monitor locking, here are things you may
want to check or try:

(1) d.mon -r monitor_name
(2) d.mon unlock=monitor_name
(3) remove the machnie_name directory under the locks directory of your grass
    installation directory. (i.e., /grass/locks/ansel in the above case)
(4) remove the .gislock (if any) in your or users' home directory

I am not sure which one will work, but it won't hurt to try all if it is not
working anyway.

The above procedures work for fifo.  For the MessQ on SUN Solaris 2.x, you need
to run the followings as root or the user who destroyed the monitor.

        ipcs -q

        This will report the ID, OWNER, and other information.  Usually there
are two in pair for GRASS Monitor processes.  You can run

        ipcrm -q ID_number

twice for each ID.  This will remove the processes of the destroyed GRASS 
monitor.

These are methods I tried if I got any monitor locking problem.  So, I can
always restore any destroyed monitor and run it again on different platforms
(Sun Solaris, Sun OS, IBM RS/6000, and Linux).  I usually don't use x0.  The
x0 seems to have many problems to me for some unknown reason.

Kang,