Factors Affecting Timestamp Values

The following factors must be considered when evaluating timestamps.

• In a multitier networked environment, there is usually no storage on the network. However, large and sophisticated server farms can employ a load balancer (for example, F5), so some data storage of the HTTP request can affect timestamp values. Since these dedicated devices are highly customized for speed, this impact is hardly noticeable.
• Content-caching devices in the environment can distort time values. When content is cached, the time values are reduced.
• For large average page sizes, it is more efficient to compress data for faster transfer.
• If the HTTP/1.1 Keep-Alive option is enabled, it can affect the rate at which multiple HTTP requests can be made.
• Application Delivery settings impact performance measurement. Is buffering turned on? I.e., does the application start transmitting when the first byte is ready, or does it wait for the entire page to first be ready?
• Chunking affects timestamp values. The answer can be delivered in one chunk, or it can be chunked and delivered on demand, such as a PDF file with byte serving.
• Client-side browser settings can affect performance and time values. For example, if caching is enabled, transfer times of pages that are containing cached content are reduced. The size of the cache is a factor.