a computer health monitor

introduction

A computer health monitor is part of system monitoring. All the systems that make up one complete running working useful computer are included.

System components to keep an eye on in one computer fall into four main categories.

  • application
  • OS (Operating System)
  • computer
  • storage

what it is

Computers were designed by man in his own image (if you look like Bertie Bassett) and are fallible. Computers fall over and die, they run amok, and they do funny things during sun spot activity (so my support engineer tells me as he hits the <CTRL, <ALT> and <DEL> keys). Users consume all system resources, give passwords to their dog and remove essential components.

When things are broken, you need to be reactive: you must find out what bit is broken and fix it or work around it immediately. When things are working, you need to be proactive: you need to know which bits of the system may break in the future and which bits of the system you can brag about in your next appraisal meeting.

Keeping an eye on system components is easy. Over the years, many people have created little UNIX utilities for looking at how a system is performing. Logging in to lots of hosts, executing all these little utilities and analysing the results is boring. Doing it regularly is tedious. Where there's boredom, there's brass. Plenty of companies sell products that will do your monitoring for you.

system monitoring
system type things to monitor description
application
  • process monitoring
  • log file scanning
  • data file monitoring
  • directory statistic
Applications include all services such as web, directory, telnet, ftp and e-mail
OS (Operating System)
  • health monitoring
  • performance management
Health checks include disk, CPU and memory use. Performance is increased by getting rid of bottlenecks.
computer
  • hardware configuration management. memory, boards, power supply.
  • fault management
  • health monitoring.
A domestic PC has sensors built into it for measuring things like temperature and memory. An enterprise PC has more sensors than a domestic PC.
storage
  • fault management
  • configuration monitoring
  • predictive analysis
This includes every storage device from a little disk drive to a great big NAS (Network Attached Storage)

what it isn't

A computer monitor. That is a display device.

A manual process. Keeping an eye on a computer's health is very dull.

where it is

Every hardware and software system needs to be monitored.

history