Answer the question
In order to leave comments, you need to log in
How to test the data aggregation system?
I ask for help with advice on how to structure testing of the data aggregation system.
Given. There is a main application on which different users perform various actions. Data on the daily activity of users is sent daily to the statistics calculation service. This service stores raw data snapshots of each user's daily activity. Then it aggregates them by day of the week and then by month in json. Graphs are then built from this data.
In this service, I see only one way to test. Manually fill in raw data for 2 months. Through excel, calculate the results of the aggregation for each day and for each month. After submitting the prepared raw data to the input, compare the output values with the data from excel.
I'm interested in how other developers test such systems and what pitfalls can be. Thanks in advance.
Answer the question
In order to leave comments, you need to log in
Manually stuffing raw data is only necessary to check for some special matches of activities or boundary ones that may not be correctly processed by the aggregation logic or to explicitly demonstrate what works as it should. Here the ex-fur-tree also will not be necessary most likely.
Moreover, it is desirable not to hammer the test data directly into the database, but to emulate sending them through the main application.
But for running through Excel, you can use real data from users.
Didn't find what you were looking for?
Ask your questionAsk a Question
731 491 924 answers to any question