Build microservices on Azure.
4:Data considerations

Polyglot persistance

Challenges

Redundancy

Data integrity

Consistency

Approaches to managing data

Embrace eventual consistency

Single source of truth

For transactions

Scheduler Agent Supervisor and
Compensating Transaction

Keep items in durable queue

Store only data that needed

Coherent and loosely coupled

Event-driven architecture

Aggregation or batching at high scale

Decouples publisher and subscriber

Application: Choosing the data stores

Delivery service

Delivery History service

Package service

Azure Redis Cache

Azure Data Lake Storage

CosmosDb

Time-series data analysis

Quick ID lookup

Fast, frequent read-write

Long-term storage

High write throughput

Simple ID queries

MongoDb API for CosmosDb