| System
Monitoring Tools
Realizing
that up-time is of extreme importance to mission critical sites, CEC
Systems has built a number of system monitoring scripts into the operating
system of each server on our network. These scripts are designed to
monitor, report, and fix various system processes according to a set
regiment of criteria. With the assistance of the CEC System's NOC
staff system administration, operating a secure, stable, and reliable
web server has never been easier. This list describes some of the
proactive monitoring that is constantly being performed on your server:
Machine
Uptime Monitor
We
have scripts that ping the servers every couple of minutes to make
sure that the boxes are accessible from within our network as well
as from outside our network. If the server cannot be pinged, the script
will notify our system engineers immediately. That way, our clients
are always confident about the performance of their server and our
network.
Web
Serving Monitor
One
of our tools monitors the Apache web server on a one minute interval
to ensure that it is up and serving pages. Every minute, we request
a page from the http daemon. If it does not respond, our script will
flag it and test it next time through. After 3 attempts the tool executes
a reload on the http daemon. Then a minute later it will request a
page again. If there is still no response, after one more attempt,
a page is sent to our support to manually look into the situation.
This script works to make sure that our clients' domains are serving
pages 24/7.
Disk
Space Monitor
Our
system also monitors disk space usage through a script that checks
the free space on our hard drives. If the available space reaches
a dangerously low level, our NOC staff would be notified immediately.
So, for example, if the log file(s) begin to grow at an alarming rate,
we would be able to prevent the log(s) from filling up the drive and
possibly bringing the server down.
Memory
Monitor
This
monitoring tool will make sure that no process takes up more than
a certain percentage of your machines' available memory. If a process
consumes more than the maximum allotment, it will be temporarily killed
off and the situation will be reported to our NOC staff.
Port
Monitor
We
actively monitor all commonly used ports, not only to make sure that
they are open, but to verify that they are actually passing data.
The script does this by asking for data. If the port does not respond,
it will be restarted by the monitoring tool.
|
|
