Real user measurement

from Wikipedia, the free encyclopedia

Real User Measurement (also briefly RUM ) is in the computer science for performance monitoring (Performance Monitoring) of web applications from the perspective of the end user ( client ).

The need for Real User Measurement is based with the so-called component myth (Engl. Component myth ). This means that in reality, performance deficiencies are often observed by the end user, although monitoring services for the individual infrastructure components do not reveal any errors. Real user measurement in performance monitoring is therefore based on the top-down principle. Only when the performance experienced by the end user on the client is ensured does the control of the performance of individual components make sense. The human response time serves as a guide value , the human reaction time to perceive delays.

Technically, the Real User Measurement with the measurement is end-to-end response times ( response time implemented) between servers and client. This can be implemented, for example, with the Open Group standard called Application Response Measurement .

Real user measurement is often part of end-to-end monitoring. This passive web monitoring is also joined by active web monitoring in the form of robot technology (also known as synthetic transaction), as well as various system monitoring components (servers, databases, virtual systems, networks, etc.).

Other terms in this context are real user monitoring or end user experience . It should be noted, however, that the end user experience also considers user behavior in addition to performance-relevant measurements. On the one hand, this data is helpful for marketing activities (keyword conversion rate ), on the other hand, in the event of an error, it can also be used to carry out an error analysis without interaction with the user (keyword error 404 ).