December 27, 2019

As a passionate developer, I love to share on technical subjects with other developers

French: Retour d’expérience Go, gRPC, Kubernetes at TF1

This talk is about software architecture and explains:

  • Why we choose Go for our technical stack, what are the benefits
  • How we used gRPC / Protobuf together
  • What are some performances issues we had in production, how we solved them
  • How we monitor our production environment with Grafana / Prometheus
  • We we deploy our microservices using Kubernetes

Software Architecture - All you need to know

This talk is about software architecture and explains:

  • What are the different architects types
  • What are their jobs everyday depending on projects workflow
  • Example of some patterns they can use
  • Which concepts every developer/architect should have in mind when making decisions

Workshop: Build a message broker using Go

This workshop goal was to learn the Go language and build a Go application from scratch within 3 hours.

The message broker has been built with:

  • Define queues and workers number per queue in a simple JSON configuration file
  • Ability to receive income messages using UDP protocol
  • Define some consumers (called processors) for each queue
  • Expose some statistics (message consumption per second, messages remaining, …)

Talk: Continuous integration/deployment using Gitlab CI

This talk was done at Eleven Labs Galatic Meetup on 11th november 2016 and was about Gitlab CI component. It is spoken in french and approach the following things:

  • Quickly: what means continuous integration and continuous deployment?
  • Why use Gitlab CI?
  • How Gitlab CI runners are working? how ot use them?
  • How to write the gitlab-ci.yml file? how does it work?
  • Some tips that can help for testing and deploying (performances, …)

Workshop: Consul - Service Discovery and Failure Detection

This workshop was about HashiCorp Consul product which aims to register your architecture services and monitor them in order to take out services that cannot be contacted anymore (because of multiple reasons: could be because there is no disk space left on the machine or an application error itself).

This workshop used the following things:

  • Docker Swarm: in order to create a little Swarm cluster which uses Consul for service discovery part
  • Docker machine: we create 3 machines for this workshop (1 consul machine and 2 node machines)
comments powered by Disqus