****************************************************
Monitor the 5.0.x Robot Dewars:

Just like monitor_dewars.com, once a page is sent, the program exits.
Somebody needs to manually go and start it again if you want to get another
page.  You start it like this:

ssh mcfuser@bl831.als.lbl.gov
monitor_robots.com >&! /dev/null &
exit

To stop it just do this:

ssh mcfuser@bl831.als.lbl.gov
killall monitor_robots.com
exit


By default:

- all three robots are monitored
- a page will be sent if any monitored robot level sensor reads > 2.1 or < 0.4
  cm for more than 10 minutes straight.
- the page "victim" is whomever is listed as "on call" in
  http://bcsb.lbl.gov/Schedules/WeeklySchedule.pdf

You can specify which robots you are interested in by naming the beamlines and
you can specify a different "victim" by mentioning them by name.

EG:

monitor_dewars.com 503 Snell >&! /dev/null &
will page Gyorgy if 5.0.3's robot misbehaves

monitor_dewars.com 501 503 >&! /dev/null & 
will page Anthony and Corie (at the moment) if either 5.0.1 or 5.0.3
misbehaves (but not 5.0.2)

monitor_dewars.com Anthony >&! /dev/null &
will page Anthony if any robot misbehaves

*******************************************
To Launch the Pager Program:

rsh mcfuser@gateway
page_when_up.com >>& /data/log/page_when_up.log &

*********************************
The sumup.com scripts are running on all beamlines:
501, 502, 503, 821, 822, 831, 1231

(It used to be necessary to "setenv beamline 822" before launching sumup.com
on that beamline.) 

Right now, all the sumup.com scripts are running on:
bl     machine  account  directory
5.0.1  bl501k2  jamesh   /home/staff/jamesh/public_html/501
5.0.2  bl502k2  jamesh   /home/staff/jamesh/public_html/502
5.0.3  bl503k2  jamesh   /home/staff/jamesh/public_html/503
8.2.1  bl821k2  jamesh   /home/staff/jamesh/public_html/821
8.2.2  bl822k2  jamesh   /home/staff/jamesh/public_html/822
8.3.1  bl831    mcfuser  /home/mcfuser/public_html
12.3.1 bl1231   sibyls   /home/sibyls/public_html

The restart procedure for the 5.0.x and 8.2.x pages is (e.g., 501):
ssh root@bl501k2
su - jamesh
cd public_html/501
../sumup.com continuous < /dev/null >&! sumup.log &
exit

The restart proceedure for 831 is:
rsh  mcfuser@gateway
killall sumup.com
cd public_html/
./sumup.com continuous >&! /dev/null &

********************************************************
Restarting the LN2 pressure Logger:

ssh root@gateway
cd /var/www/html/pressure
./lN2_mon.com >& /dev/null &
**************************************
Restarting the axis2 Web Camera Server:

~jamesh/beamline/http_bridge.tcl axis2 80 12345 >& /dev/null  &

****************************************************

HOW TO CHANGE GATEWAY MACHINE FROM, say crush8 to crush9

swap the CAT5 line from the outside world hub
from ethernet card in crush8 to crush9

logon to crush9 as root
in /root is a file called "become_server.com"
open the file, then cut and paste commands (or just run it, it's now an executable)

Reboot the computer 

**************************
back way into BL831 if gateway is down is through bcsb-staff-1.als.lbl.gov,
then ssh  to bl831a
bcsb-gateway is currently offline.

*******************************************************
in /root of dataserver is "things_to_restart.com"
*****************************************************
NFS hang problem symtoms are:

  adxv is not updating
  "index" hangs (so will "df" and "ps")

Solution:
  ssh root@graphics1
  mount /data

Double-mounts are allowed, and pretty much safe as an interim solution.  To do
a clean unmount, you need to issue a "Ctrl-C" or "kill" to everything trying
to use /data, and then execute several "umount -f /data" commands until it no
longer gives you an error.  Then you can "mount /data" again.

More complicated, so my middle-of-the-night advice is just to "mount /data"

 From: JMHolton@lbl.gov
 Date: Saturday, July 24, 2004 0:19 am
 Subject: graphics1 "froze" again
******************************************************************
Here's how to restart detectors without killing and restarting the WindowsNT "Romote
Detector OP"  (copied from nuke):

mcfuser@graphics1:/home/mcfuser 2% foreach module ( detector1 detector2
detector3 detector4 )
foreach?     # send the reset signal
foreach?     echo -n "$module restart "
foreach?     echo "restart" | sock_exchange.tcl $module 8038 1
foreach?     echo ""
foreach? end
*****************************************************************