We manage what we measure. Measuring Web application performance ensures we know what our customers experience and stay informed as change happens. This section provides answers to common questions about Web monitoring for beginners.
What is Web monitoring?
Web monitoring provides Web site and application owners with information that’s essential to optimize Web application and service performance:
- Alerts – that notify operations about specific system errors and related information.
- Trends – providing insight into performance data consistently over time.
- Analytics – to determine the root cause of issues, especially among the many stakeholders typically involved in delivering a Web application or service including end users, browser and mobile device vendors, developers, and various network and Cloud service providers.
Organizations proactively monitor because:
- User experience has increasingly become a competitive advantage
- Web and mobile applications getting richer, more complex
- Many stakeholders in the Web application delivery chain and problem diagnosis
- Less control over content delivered from third party sources
- More capabilities residing in many possible browsers and mobile devices
- It’s nearly impossible to know what your customers experience without monitoring.
The “SLA paradox”: uptime versus response time
Uptime is only half the picture. You can be up but still slow. More importantly, when is it slow? Is it during peak hours? In which case, you could be giving a significant portion of your visitors poor service even though you appear to be up all the time. We refer to this as the “SLA Paradox.” The only way to mitigate it is to monitor performance consistently over time.
What is response time?
Website and Web application response time is made up of four main components:
- Internet Lookup and Routing Time – the time it takes to establish connection between your server and the end user’s computer.
- Server Wait Time – the “time to first byte”, occurs once the connection is established and the server gets prepared to then serve content.
- Content Loading Time – the time between when the first byte is sent until the user receives all content.
- Page Rendering Time – the time it takes to render content in a user’s browser.
What slows Web performance down?
Possible causes are endless. Some of the most common ones include:
- Database/Back-End Connections –heavy reliance on databases and various back-end systems
- Caching – improperly-configured front end cache systems
- Heavy Content – large files and rich media
- Third Party Content – embedding external content and links ties your site to theirs.
How does monitoring improve Web performance?
The devil is in the details. Monitoring allows you to see not only the details but also how they trend over time. Consistent, controlled measurements and trends reduce time to discover bottlenecks and root cause, provide critical insight into solutions, and hold stakeholders accountable for their service commitments.
How is response time related to maximum load capacity?
Response time is a key indicator of maximum load capacity
. Low, stable response times at low traffic volumes may indicate more capacity is available. But if response times are high, or if they are low with high variations, this indicates your Web application will not handle more traffic well. The only way to know is to investigate response time patterns.
Apica Global Monitoring Network