PIDFILE cleanup broken in sample monitor?

I came across the above error after deploying an EAV monitor in LTM 10.2.3

After debugging and stepping through the script we were able to confirm that it never reaches the PIDFILE cleanup line on successful execution:

mark node UP if expected response was received

if [ $? -eq 0 ]

then

echo "UP"

fi

rm -f $PIDFILE

This appears to be 'normal' as result of any STDOUT generated by the script causing bigd to terminate the EAV process, as described in http://support.f5.com/kb/en-us/solu...l7444.html

In order to render the logged warning useful and only triggered under the condition where the script failed to exit, shouldn't we be removing the pidfile AFTER the test and BEFORE any output?

i.e.

mark node UP if expected response was received

if [ $? -eq 0 ]

then

rm -f $PIDFILE

echo "UP"

fi

rm -f $PIDFILE

or even cleanup before evaluation of the test error code (requires setting a var for the return code):

send request & check for expected response

curl -fNs http://${NODE}:${PORT}${URI} -d "${DATA}" | grep -i "${RECV}" 2>&1 > /dev/null

RC=$?

remove the pidfile, since the monitor executed

rm -f $PIDFILE

mark node UP if expected response was received

if [ $RC -eq 0 ]

then

echo "UP"

fi

exit

Is anyone else getting the same results or am I completely missing something here?

See also https://devcentral.f5.com/wiki/AdvDesignConfig.HTTPMonitor_cURL_BasicPOST.ashx

Matt