After an alert is generated, the most effective way to understand how one specific application is behaving is by looking to a graphic with all data.
It is very important to find the answer to the following questions:
- if a performance degradation are all users experiencing the problem? or just a part of the users?
- how bad is the problem? are the maximum times near 10s or 5s?
- is the degradation near one specific value or near two specific values?
- where are the majority of the times of the transaction?
In a graphic that shows all data, without any grouping of the data, it is possible to answer this questions. In RUM there is not that possibility and the graphic more close to this is the "Triage Raw Data Report".
In the online user manual customer have found the following information:
"Note: For real user data, each colored square represents the average response time for the defined time range and is a good approximation of the raw data".
The customer needs the raw data and the only way we are getting this information is by exporting the data to another application where they can generate a graphic where each point represents one transaction.
They need a report like "Triage Raw Data Report", but where one point represents one transaction and not an average of transactions.
To have a report which represents one transaction not the average of transactions for RUM.