A breif about advanced skills in backend programming.

1. Context

Backend skills can be divided into two levels:

• Basic level: focuses on connecting and interacting with infrastructure (Database, Redis, Kafka) using protocols like HTTP, gRPC, GraphQL, WebSocket, UDP,…

• Advanced level: deals with concurrency programming, handling multiple requests and processes through inter-process communication and managing concurrent database/Redis access.

In this blog, we will diccuss about concurrency programming in Backend Programming with main concepts, using Golang language.

2. Advanced Backend Programming with Go

2.1. When to use

When you need to handle multiple background tasks that can run independently, such as:

• Pushing log to analytics dashboard: Google Analytics, Kibana,…
• Cache warming and maintenance jobs: reconcilation, alerting,…
• Create multiple workers (worker pool) to split jobs independently for a long-running jobs.

Notes: All of the tasks is not depend to the main business process => if you call this functions synchoronously, it will cause the high latency for your service but do not necessary.

=> Parallel programming enables significant performance improvements by executing these tasks concurrently, reducing overall latency and improving system throughput.

2.2. Issues

• Race condition: and it causes crash services due to the conflict when multiple processes access to the same memory address at the same time.

• Data inconsistency: when you do not follow ACID principle while updating database.

2.3. Goroutines

• Each goroutine you can view as an process to run functions parallelly with each other.

• Main function is a goroutine.

2.4. Channel

• Channel is a queue for multiple goroutines (process) to communicate with others.

• Implemented by pubsub mechanism.

2.5. Wait Group

• Used to wait N goroutines to complete their job.

2.6. Mutex

• Used to lock the process in 1 goroutine.