35 lines
1.7 KiB
Text
35 lines
1.7 KiB
Text
===========================================================================
|
|
$NetBSD: MESSAGE,v 1.2 2002/09/24 12:30:30 wiz Exp $
|
|
|
|
QUICK START GUIDE:
|
|
|
|
Change your directory to the now installation base, eg:
|
|
${NOW_ROOT}/bin. Run ./glupart install as root.
|
|
|
|
You must initialize the partition management system to
|
|
understand the machines in your cluster. The tool to use is
|
|
"glupart". You will need to do a "glupart machine <name> new" for
|
|
each new machine. Multiple machines can be added at a time using a
|
|
comma-separated list of names. You can also do "glupart -help" for
|
|
more detailed information.
|
|
|
|
To run GLUnix, you need a single machine which is able to do a root
|
|
rsh (or ssh or kerberized rsh) without a password to all of the
|
|
machines you want to run GLUnix on. Call this trusted machine the
|
|
"master" machine. The scripts that get GLUnix running initially (the
|
|
run_glunix_* scripts) need to run an rsh command on each node in order
|
|
to set up GLUnix. Log on to the master machine and run "run_glunix
|
|
-start <nodes>" where "<nodes>" is the list of nodes you want to run
|
|
GLUnix on: generally a comma-separated list of machines. (You can do
|
|
"run_glunix -help" to see all the options.) If the names of some of
|
|
the machines are of the form of some base string followed by
|
|
consecutive numbers (our machines are u0, u1, u2, u3, u4), then these
|
|
machines can be abbreviated to the run_glunix command as "u0..u4".
|
|
Here is a sample run_glunix command:
|
|
|
|
run_glunix -start thunk,sparc,u0..u110,pc0..pc9,dawn0..dawn30,cayuse
|
|
|
|
Once the run_glunix command has successfully completed, you can run
|
|
"glustat" to find out how many nodes came up.
|
|
|
|
===========================================================================
|