Chandani Patel

Hi,

Thanks for reading my article.

Here is my response to your queries.

1. When reports needs to aggregate data ?

Aggregation of data depends on what reports you are generating,
if you are generating the validation reports you should not aggregate data because these reports are used in quality check and certification of the product.

If you are monitoring the infrastructure then you need to consider multiple aspects like service health, computing resources, response codes, spikes in load, etc. which needs different aggregations like average, sum, min, max
I always prefer the raw data formats and aggregate on need bases using tools and we can define the reporting based on our need. There are many opensource tools which allow us to do this within a few clicks if we can transfer data to database eg. grafana, BIRT, Metabase

2. When customer requirement changes and it affect the reports should we change all affected services?

As per microservices architecture, any change in single service should not affect any other services, you should try to divide services in small chunks and you can avoid multiple changes.
You can also define templates for data (eg. key, value) and use those templates to generate reports so you need not change anything on reporting when requirement changes

3. Should we require to foresee future reporting requirements when designing microservices?

Yes, Reporting is a very important part of any successful system. It gives us insights for growth and system and prevents multiple failures.
IMO while designing we should think about Reports of Infrastructure, Development and Validation Monitoring, and few business reports which can bring first level insights of system.

Thank you, Srikanth.