yet another run-across-cluster script
Joe Landman
landman at scalableinformatics.com
Fri Mar 28 09:39:55 PST 2003
Hi folks:
I wanted to give a quick pointer to a simple cluster tool I had built
a while ago (that I seem to use everywhere). It is
yet-another-run-a-command across all the nodes, though it uses ssh,
doesn't require daemons (other than SSH), doesn't send passwords or
usernames, and so forth. No, it doesn't do parallel dispatch (yet).
You can pick it up at
http://scalableinformatics.com/downloads/all-0.25.tar.gz . Usage is
quite simple:
[landman at squash.scalableinformatics.com:~/all-1.0]
69 >./all uptime
username:
password:
----- machine =crunch:
.
.
.
.
crunch: out= 11:03am up 3 days, 18:03, 4 users, load average:
0.02, 0.01, 0.00
, error=, exit=0
----- machine =squash:
.
.
.
.
squash: out= 11:04am up 8 days, 27 min, 2 users, load
average: 0.48, 0.43, 0.33
, error=, exit=0
----- machine =stuff:
.
.
.
.
stuff: out=blank, error=blank, exit=blank
-------------------
failed machines=stuff
(minimal) Documentation included. No extra daemons required (uses
sshd).
Joe
--
Joseph Landman, Ph.D
Scalable Informatics LLC,
email: landman at scalableinformatics.com
web : http://scalableinformatics.com
phone: +1 734 612 4615
More information about the Beowulf
mailing list