So a light bulb went off and I decided to look into service monitoring tools or scripts that would restart a service if it crashed. I found Monit (GNU General Public License v3).
I have had it installed for only a day so far but ntop has stayed alive for me. So far so good. below are the steps I did to install it. Some might find the ntop config for the Services section useful. I am not an expert in Monit so I am sure there may better configurations for ntop then the one I have below. Feel free to share yours so we can see.
1) Install Monit
fedora: # yum monit
debian: # sudo apt-get install monit
2) Next, edit the config file
/etc/monit/monitrc
on Fedora
/etc/monit.conf
--------------------------------
### Monit control file /etc/monit.conf
set daemon 120 # Poll in 2-minute intervals
set logfile syslog facility LOG_daemon # Default facility is LOG_USER
set mailserver mail.foo.bar # Default smtp server is localhost
set alert sysadm@foo.bar # Alert system admin on any event
set httpd port 2812 address localhost
allow localhost
allow admin:monit
## Services
### NTOP on NST 2.13.0 ###
check process ntop with pidfile /var/run/ntop.pid
start program = "/etc/init.d/ntop start" with timeout 60 seconds
stop program = "/etc/init.d/ntop stop"
if failed port 3000
with timeout 15 seconds
then restart
if 3 restarts within 5 cycles then timeout
group server
--------------------------------
3) Save the monit.conf file
4) Now check the syntax with:
moint -t
(fix the syntax if you need to)
http://mmonit.com/monit/documentation/monit.html
5) Make sure Ntop is running (ports 3000 and 3001 for my installation); I use
# nmap localhost
6) Start Monit
# /etc/init.d/monit start
7) Login into the web admin http://localhost:2812
with u:admin p:monit as configured in the monit.conf
8) Enable monitoring of ntp by clicking on 'ntop' or navigate to
http://localhost:2812/ntop
on the bottom of the page click the "Enable monitoring" button.
It will take some time but the "Monitoring status" go into "monitored" mode.
