Deploy and Scale Machine Learning Models With Kubernetes

Meïssa B.C MBAYE

1

Who Am I ?

Software & DevOps Engineer at @Baamtu
FullStack Developer
CyberSecurity & DevSecOps Enthusiast
Open Source & Tech Lover

2

Summary

📝 Intro
🎨 VM vs Container
🧑‍💻 What is Docker ?
🤹 What is K8S ?
🎥 K8S Architecture
📤 K8S Terminology
🛠 Demo

3

Introduction

Scaling and industrialize AI development/deployment

4

Introduction

As enterprises increase their use of artificial intelligence (AI), machine learning (ML), and deep learning (DL), a critical question arises: How can they scale and industrialize ML development ?

5

VM vs Container

Virtualization and Containerization

6

VM vs Container

What is Virtualization ?

"In computing, virtualization or virtualisation is the act of creating a virtual version of something at the same abstraction level[…]."

Wikipedia.

7

VM vs Container

8

Docker

What is exactly Docker ?

9

Docker

Is a containerization technology that allows the creation and use of Linux® containers.
Offers several advantages such as :

Modularity
Layers and image version control
Rollback
Rapid deployment

10

Docker Workflow

11

Docker is good but…

12

Kubernetes

What is exactly K8S ?

13

Kubernetes | k8s | kube

Is an open source container orchestration platform or framework.
Automates many of the manual processes involved in :

Deploying
Managing
Scaling containerized applications.

14

K8S Architecture

How Kubernetes Works ?

15

K8S Architecture

16

K8S Terminology

Presentation of some kubernetes components

17

K8S Terminology


`Cluster`	A set of worker machines, called nodes, that run containerized applications.
`Deployment`	An API object that manages a replicated application, typically by running Pods with no local state.
`Pod`	The smallest and simplest Kubernetes object. A Pod represents a set of running containers on your cluster.
`Node`	A node is a worker machine in Kubernetes.
`Service`	An abstract way to expose an application running on a set of Pods as a network service.
`Kubectl`	Command line tool for communicating with a Kubernetes cluster’s control plane, using the Kubernetes API.

18

Demo

Demo time

19

Demo

20

Thank You

😎😎😎

21

Resources

22