Concurrency in Go

Personal notes on the book "Concurrency in Go" by Katherine Cox-Buday. Reference: Katherine Cox-Buday. Concurrency in Go: Tools and Techniques for Developers. O'Reilly Media, 2017.

Table of Contents

• Concurrency Semantics
• Communicating Sequential Processes
• Concurrency Building Blocks (in Go)
• Concurrency Patterns (in Go)
• Concurrency at Scale
• Goroutines and the Go Runtime

Concurrency Semantics

• "Concurrency", "asynchronous", "parallel", and "threaded" are distinct concepts
• A process is considered concurrent if it occurs simultaneously with one or more other processes
• Amdahl's Law models potential performance gains from implementing solutions in a parallel manner
  • It helps determine whether parallelization is an effective strategy for addressing performance concerns in a system
  • Gains are limited by the portion of the program that must be executed sequentially
  • Evaluate whether significant performance improvements can be achieved by simply increasing the number of cores available to your program
  • Consider if the problem can be reduced to how to combine and store results obtained in parallel
• If your problem is embarrassingly parallel, design your application to scale horizontally whenever possible
• Cloud computing introduces challenges such as:
  • Provisioning resources
  • Communication between machine instances
  • Aggregating and storing results
  • Modeling code for concurrent execution
• Web scale enables properties like:
  • Rolling upgrades
  • Elastic, horizontally scalable architecture
  • Geographic distribution
• These properties introduce new levels of complexity regarding comprehension and fault tolerance

Concurrency is Hard

Communicating Sequential Processes

Concurrency Building Blocks (in Go)

• sync package: handles Go's memory access synchronization

Concurrency Patterns (in Go)

Concurrency at Scale

Goroutines and the Go Runtime