Gain insights into adopting Elixir for web apps. Explore staffing, code management, functional transition, distributed Elixir, deployment, performance assessment, and more, for a seamless transition.

adelixir.tar.gz

BeliefStructure

BeliefStructure-credo

BeliefStructure-dialyzer

BeliefStructure-inchex

BeliefStructure-coveralls

BeliefStructure-bureacrat

BeliefStructure-bh

BeliefStructure-Aliases

BeliefStructure-format

BeliefStructure-dialyzer-wrong

BeliefStructure-terraform

BeliefStructure-terraform-copy

Umbrella Projects

counter

path_alllocator

path_alllocator-copy

EchoServer-Java

Distillery

observer-copy

formatter

formatter-fixed

sasl_reports

load-testing

profiling

benchee

ecto-log-copy-copy

terraform

observer-real

terraform-postman

nifs

BeliefStructure-coveralls-copy

BeliefStructure-inchexB

Elixir is a relatively new framework but shows rapid growth due to its flexibility as a functional, concurrent, distributed programming language.

This course is a practical guide to adopting an Elixir-based technology stack for distributed web applications. Engineering managers will find helpful guides on upskilling and cross-skilling their teams, and technical leads will grasp the meaningful differences between Elixir and traditional stacks. The course moves from high-level concepts such as staffing, code management, and legacy systems to developing your Elixir environment with a walkthrough of making the functional transition, distributed Elixir, and integrating external code. You’ll wrap up with deployment and maintenance functions, assessing performance, and finalizing the application.

By the end of this course, engineering managers will be prepared to tackle the unique challenge of migrating web applications to Elixir.

Adopting Elixir for Software Engineering Leads & Managers

# Some useful strategies

In this chapter, we explored three strategies for integrating external code:

* **Strategy 1:** 

   Native implemented functions (NIFs) allow developers to load code into the same memory space address as the Erlang VM. It is the most performant option but also potentially unstable and insecure. A crash on a NIF can bring the whole node down, and long-running calls can cause the VM to misbehave. However, dirty schedulers can mitigate some of the disadvantages.

* **Strategy 2:** 

  Ports run on the same machine as the Erlang VM node but in a different process. They use I/O for communication. Ports are generally the best option to consider for integration, as they does not contain any of the safety implications behind NIFs nor the overhead behind the Erlang distribution protocol.

* **Strategy 3:**
 
  The Erlang distribution protocol allows other platforms to piggyback on Erlang VM’s ability to communicate between nodes to integrate other languages. It has the same downsides as using the Erlang distribution for heterogeneous systems, since it is effectively a heterogeneous system.

That wraps up Part II: Development. We explored building clean code based on clean abstractions, as well as how to instrument and measure results so it stays beautiful and easy to implement. Then, we worked through integration strategies for other nodes and external libraries.


Get a summary of how Elixir can integrate with the external code in this lesson.

Wrapping Up

Summary of Integrating with External Code

Introduction

Three Adoption Stories

Concept: Team Building

Concept: Ensuring Code Consistency

Concept: Legacy Systems and Dependencies

Development: Making the Functional Transition

Development - Distributed Elixir

Development - Integrating with External Code

Production - Coordinating Deployments

Production - Metrics and Performance Expectations

Production - Making Your App Production Ready

Conclusion

Appendix

Summary of Integrating with External Code

Some useful strategies