Kafka is best fit here. Below are the advantages with Kafka ACLs (Security), Schema (protobuf), Scale, Consumer driven and No single point of failure.
Operational complexity is manageable with open source monitoring tools.
4 upvotes·552.7K views
Naresh Kancharla
Staff Engineer at Nutanix
I use gRPC because it is based on http/2 and provides advantages when used at scale. With protobuff grpc will be more efficient and easy to integrate. Grpc can give efficient way to talk to server it can open persistent connections and also asynchronous communication between client and server.
3 upvotes·26.4K views
I use MariaDB because its open source and relational database. Out of open source relational database MariaDB stands out providing scalability and availability. Other relational databases needs DBA personas but MariaDB made this easy.
3 upvotes·2K views
2 upvotes·28.9K views
I use Envoy because it supports both http and http/2. It supports grpc and rest. It can even translates REST to Grpc. Best functionality i like from Envoy is it generates metrics which can be imported by Prometheus.
2 upvotes·12.4K views
I use GraphQL because it is driven by client and api user defines the input schema which reduces number of end points on backend. End points will be delivered quick. It gives response as per DB schema but this can customised.
2 upvotes·258 views
I use MongoDB because its document store with json and easy to store and query. Less learning curve. Give lot of query capabilities and next door to relational database.
1 upvote·111 views
It depends on your data and read/write pattern. Do have read operations on multi dimension where clauses? Do you have update operation on any column ? Is your data relational ? I use RDS if above are true as slice and dice on the data is easy. Where as Dynamo is key value or Json like. Schema less.
1 upvote·96 views