Unix Performance Measurement
| Unix Performance Measurement |
|
There are two basic type of measurement for Unix, real-time i.e. a snapshot of what is happening at that moment or logged. Both have their own uses. |
| Logged Data |
|
This primarily to give a history of the systems performance over time and is used for capacity planning purposes. It is also very useful for highlighting a gradual tail-off in performance due to addtional users begin added for example. Most logged data is really only real time data collected at regular intervals. Although DBAudit is primarily a database monitoring tool it can be extended to monitor any Unix or Business metric. |
| Real Time Data |
|
This is a quick snap-shot of what is happening now. In most cases nothing meaningful can be deduced from this, what is required is a number of snap-shots run in quick succession. This will identfy an immediate problem. |
| Unix Commands |
|
Due to the number of Unix's available the term standard is very misleading when discussing commands. Each vendor provides their own version of the tools and while the data might look the same between platforms it always worth investigating what is really being measured and is it the same on two different platforms. |