When working with Spring Data for file based integration, I was wondering how the application is handling the file processing. We have seen file descriptors left opened after the application run for a while. and the application eventually stalled because an IO Exception being through out for the files opened by the application have reached its maximum.
The following script help me to monitor the file descriptor opened in the Linux server:
#!/bin/sh
#
#published GNU GPL3.0
#please change /foo/bar to the folder you want to mornitor
PID=`ps -eaf | grep tomcat | grep -v grep |grep org.apache.catalina.startup.Bootstrap | awk '{print $2}'`
if [[ "" != "$PID" ]]; then
DATE=`date '+%Y-%m-%d %H:%M:%S'`
echo "PID $PID $DATE">>lsof.log
/usr/sbin/lsof -a -p $PID|grep /foo/bar >>lsof.log
fi