Call hciprocstaus from script

[RICK L. PRITCHETT]

I am trying to call hciprocstatus in an alert using a script.  Below I believe I am setting up the env variables correctly.

#!/usr/bin/bash

#set hciRoot /hci/cis19.1/integrator

eval $HCIROOT/sbin/hcisetenv -root ksh > /dev/null

site=$1

process=$2

site=basename $site

#echo $site

eval $HCIROOT/sbin/hcisetenv -site ksh $site > /dev/null

showroot

/hci/cis19.1/integrator/scripts/alert_scripts/start_process.tcl $site $process

 

start of proc called.  I get this in the log file:  Unable to contact hcimonitord: timed out

start_process.tcl

#!/usr/bin/env hcitcl

set site [lindex $argv 0]

set process [lindex $argv 1]

append filename_cmd3 $process _cmd_output3.txt

hciprocstatus -p %P -i > /hci/cis19.1/integrator/scripts/alert_scripts/%P_cmd_output3.txt

to test it I have an alert that is set up as an exec that does return the output

hciprocstatus -p %P -i > /hci/cis19.1/integrator/scripts/alert_scripts/%P_cmd_output.txt

 

any ideas?

 

 

Topic

[Peter Heggie]

It looks real close to what we do.

We do a setroot <path to “integrator” folder>

Then the setsite.

Then issue the start process command.

Then go into a loop and wait for the status to be running:

# wait for process to start
twait=0
sprocess=printf "%.15s" "$process"
tstat=hciprocstatus -p $process |grep $sprocess | awk '{ print $2 }'
while [ “$tstat” != “running” ]
do
sleep 2 ; twait=expr $twait + 2
if [ $twait -gt $waitseconds ]; then break ; fi
tstat=hciprocstatus -p $process |grep $sprocess | awk '{ print $2 }'
done

# did we ever get a status?
if [ “$tstat” = “running” ]; then
out=”started $process”
rcode=0
else
if [ -z “$out” ] ; then
out=”hciconnstatus | grep $sprocess error starting $process”
fi
rcode=8
fi

Peter Heggie

[RICK L. PRITCHETT]

I tried part of your code and I am getting an error on the -p argument.  I have tried encapsulating the hciprocstatus -p $process in single quotes, double quotes and the different brackets.

tstat=hciprocstatus -p $process |grep $sprocess | awk ‘{ print $2 }’

/hci/cis19.1/integrator/scripts/alert_scripts/setsite_alerts.sh: line 20: -p: command not found

[RICK L. PRITCHETT]

Ok, I re-started the daemons and now get a time-out.  Unable to contact hcimonitord: timed out

I noticed the lock manager restarted with the same PID.  Could this be an issue?

[Author]

Replies