Performance Testing Your Website
To ensure your website is serving your visitors in a timely fashion, one
must run routine tests against various web pages, testing for responsiveness.
The key factors in testing responsiveness are identifying problem pages, and
monitoring key pages for trends.
Identifying Problem Pages
Identifying problem pages is usually an easy task. During the development,
pages that load in a noticeable amount of time (over 1 or 2 seconds)
should be noted for monitoring. After development is complete, during the
testing phase, the notably slow pages should be monitored over time as the
pages are optimized.
Before code or database optimization is performed, a baseline should be
established. With a website monitoring tool, such as WebWatchBot, monitor
the page(s) every minute by loading the entire page (not a ping, but a
full page download). After one hour, note the average response time –
this will be the baseline average response time.
During code or database optimization, continue to monitor the page(s)
every minute. Is there a change in response time? Using WebWatchBot,
check the response time chart for a change in response time –
change the chart options so that data is not grouped together.
Continue the process of tweaking the code and database stored procedures
and queries until the response time is at an acceptable level.
Monitoring Key Pages
After website development is complete and the website is in a production
environment, it is critical to continually monitor your key pages for
responsiveness. With WebWatchBot, create watch items for your home page,
your about us page, and other pages of key importance. The frequency of
monitoring should be at most every 2 minutes, which will over time
collect a considerable amount of data.
After one week of data collection, you can chart response time with the
Response Time Chart for trends – change the chart options so that data is
grouped and the trend checkbox is checked. View the chart with data
grouped "daily by hour" to identify potential problem times during the
day. View the chart with data grouped "month by day of week" to
identify potential problem days of the week.
Over time, trends will appear and solidify a case for returning to
problem pages on your website for further optimization.
