A very wise man in a movie once put it, “It’s good for a man to know his limitations.” When it comes to a computer system, it’s essential.

Understanding your capacity is a healthy part of running a business. Any business must understand what it needs in terms of its personnel (and the resources to support those personnel) to support and sustain itself.

A technology business must take this understanding further in terms of its networking, hardware, and software infrastructure.

There are several levels of understanding involved here, namely:

  • System
  • Machine
  • Application

 
Obtaining system and machine level capacity understanding is in some ways the simplest calculation. This can be taken as a mere sum and average of the following consumption/utilization metrics used for detecting the capacity of a single machine, namely:

  • Memory
  • CPU
  • Disk
  • Network Bandwidth

These numbers are vital, and if they’re not being analyzed regularly at a machine and system level, then the system is exposed. Yet, these numbers alone don’t paint the complete picture.

To obtain the full insight, you must dig deeper into the application level, particularly into the “hot spots” of the process. A “hot spot” is where the most common processing occurs. Collecting metrics at the hot spots is how to fully gauge a system’s capacity. Such information may include any or all (or more) of the following:

  • Total Processing Time
  • Disk Processing Time
  • Network Processing Time
  • Total Data Volume
  • Disk Data Volume
  • Processing Entity Volume

By collecting the vital hot spot metrics across a system’s applications, you will more accurately determine the capacity of a system. Consequently, this will help in terms of system maintenance and growth of the system. In addition, these metrics will help predict how well the system can handle downtime (expected or not), which inevitably occurs to even the best of systems.

–J. Scotte Barkan, CTO & Co-founder of Catchpoint