It is important to understand your application's spike hours and then design your performance tests based on that data.

For example: a 'spike hour' may occur when a ticketing website sees a sudden rush in activity a few moments after tickets for a new events become available at 8 a.m. every Friday.

Spike hours are an important dataset, and you should ensure that your server configuration can manage them efficiently. Some key points worth considering while investigating the spike hours associated with your website or application during performance tests include:

• Knowing when your application will receive a sudden increase in activity (Spike Hours).
• Understanding the transaction load, the user load, other key site functionalities, user types that create the heavy load, any key geographies producing high volume traffic, etc.
• Knowing the business flows that show the most activity during the peak hours.
• Knowing the user types (admin/managers/end users etc.) that show most activity during the peak hours.
• Being aware of all the third parties involved in your application - and letting these stakeholders know about the peak hour expectations.
• Knowing the geography of the end users (the regions from where most requests are coming) and, if possible, positioning your application servers in locations that can cater to your key demographics. This will help end users experience increased responsiveness and more reduced load time.

Click here to go to the next page.