Slide 1

Slide 1 text

Cloud and Hybrid Developer Environments How virtualization transforms your IDEs for embedded projects Oleg Nenashev, PhD Independent Consultant QA&Test 2025 Bilbao

Slide 2

Slide 2 text

2

Slide 3

Slide 3 text

Disclaimer ● I am at this conference as independent, #OpenToWork ● Opinions are my own ● I am not talking about any my past or current companies speakerdeck.com/onenashev

Slide 4

Slide 4 text

> whoami @oleg_nenashev oleg-nenashev Dr. Nenashev / Mr. Jenkins Community Builder, Dev Tools DevRel & DevOps Consultant #RussiansAgainstPutin #StandWithUkraine

Slide 5

Slide 5 text

@oleg_nenashev oleg-nenashev > whoami --automotive --embedded --hardware 5 CloudBees Embedded & Automotive Day in Stuttgart LibreCores CI

Slide 6

Slide 6 text

IEEE Spectrum Research among members 6 https://spectrum.ieee.org/top-progra mming-languages-2021 Today

Slide 7

Slide 7 text

Today 7 Most of demos are for the Testcontainers for C/C++ project - https://github.com/testcontainers/testcontainers-native Not an embedded project, but created for Embedded/Automotive projects

Slide 8

Slide 8 text

Embedded Developers Can Have Nice Things 8

Slide 9

Slide 9 text

My Previous Presentations 9 ● Developer Tools & Containers ● Jenkins, Docker, WireMock, Testcontainers

Slide 10

Slide 10 text

Today ● Hybrid and Remote Developer Environments ● Dev Containers ● Cloud Developer Environments (CDEs) ● Brief: DIY CDEs Demos - coffee breaks 10

Slide 11

Slide 11 text

Is your laptop enough to work on the main project? 11

Slide 12

Slide 12 text

12

Slide 13

Slide 13 text

Your DevEnv is already elsewhere 13

Slide 14

Slide 14 text

What does engineer use daily? ● CI Systems ● Issue Trackers ● Chats ● Zoom / Google Meet / … ● Email ● Calendar ● … 14

Slide 15

Slide 15 text

CloudBees Office, Neuchatel, ~2015

Slide 16

Slide 16 text

Local Development 16 In Containers

Slide 17

Slide 17 text

Why containers? ● Configuration Management as Code ● Reusability ● Fast provisioning ● Clean environments ● Disposability ● Clustering

Slide 18

Slide 18 text

My Previous Presentations 18 ● Developer Tools & Containers ● Jenkins, Docker, WireMock, Testcontainers

Slide 19

Slide 19 text

Docker in Embedded Prototypes, Dev Tools Test Infrastructures, CI/CD Docker Swarm Cluster on Raspberry Pi

Slide 20

Slide 20 text

20 WireMock for Google Test / C++ github.com/testcontainers/te stcontainers-c/tree/main/de mo/google-test

Slide 21

Slide 21 text

Containerized IDEs 21 https://code.visualstudio.com/docs/ devcontainers/containers

Slide 22

Slide 22 text

Local Dev Environments in containers, too 22 Local Container Engine or Cloud

Slide 23

Slide 23 text

23 + IDE Integrations https://containers.dev/

Slide 24

Slide 24 text

Dev Containers 101 24 https://hub.docker.com/r/microsoft/devcontainers-cpp

Slide 25

Slide 25 text

Dev Containers 101 .devcontainer/ devcontainer.json - Configuration JSON Docker image, container parameters, IDE plugins containers.dev/ implementors/spec/ 25

Slide 26

Slide 26 text

Quick Start 1. Clone https://github.com/microsoft/vscode-remote-try-cpp 2. Open the project in VSCode 3. Install the Dev Containers extension 4. Reopen the project in a container 26

Slide 27

Slide 27 text

27 Sample: https://github.com/microsoft/vscode-remote-try-cpp

Slide 28

Slide 28 text

28 Sample: https://github.com/microsoft/vscode-remote-try-cpp 1 4 3 5 1. Environment runs in a container 2. The workspace is mapped 3. Tools are accessible 4. IDE plugins, including Git, Intellisense & Co 5. AI Agent works, too

Slide 29

Slide 29 text

Dev Containers are not just simple containers 29

Slide 30

Slide 30 text

Dev Containers 1. Stock images 2. Single container From image registry or Dockerfile 30 containers.dev/guide/dockerfile

Slide 31

Slide 31 text

Dev Containers 1. Stock images 2. Single container 3. Multi-container systems 31 containers.dev/guide/dockerfile#docker-compose

Slide 32

Slide 32 text

Dev Containers for your dev environments Supported in: ● Visual Studio Code (link) ● Visual Studio (link) ● IntelliJ Platform IDEs (link) ● Cloud IDEs and Services ● CI/CD Systems 32 https://containers.dev/ Your developer tools and environments, packaged as containers. With proper developer experience

Slide 33

Slide 33 text

Add-ons ● Templates 33 https://containers.dev/templates

Slide 34

Slide 34 text

C++ Template https://github. com/devcont ainers/templa tes/tree/main/ src/cpp 34

Slide 35

Slide 35 text

Add-ons ● Templates ● Features 35 https://containers.dev/features

Slide 36

Slide 36 text

Dev Containers 36 https://containers.dev/features

Slide 37

Slide 37 text

Add-ons ● Templates ● Features ● IDE Plugins 37

Slide 38

Slide 38 text

38 Example - CMake in a Dev Container https://github.com/testcontainers/ testcontainers-native

Slide 39

Slide 39 text

39 Example - GoogleTest in a Dev Container https://github.com/testcontainers/ testcontainers-native

Slide 40

Slide 40 text

Dev Containers Reuse 40 Source: https://containers.dev/overview#De velopment-vs-production

Slide 41

Slide 41 text

Reuse Dev Containers can be shared as a part of your project Dev Containers can be reused in CI/CD pipelines 41 https://github.com/testcontainers/testcontainers-native/blob/main/.github/workflows/ci.yml

Slide 42

Slide 42 text

FPGAs and Simulators 42

Slide 43

Slide 43 text

Testing with hardware VM, container or bare-metal Power Reset when it hangs Deploy bit-files, Blasting cables Debugging cables, firmware upload We don’t change them, huh? Your Developer Machine Drivers Tools Peripherals Attached peripherals Runtime

Slide 44

Slide 44 text

Docker & FPGA - HOWTO 44 ● USB/Ethernet Ports can be exposed to a container ● Docker image can include tools and drivers ● https://github.com/halfmanhalftaco/fpga-docker ● https://www.starwaredesign.com/index.php/blog/ 64-fpga-meets-devops-xilinx-vivado-and-jenkins-with-docker

Slide 45

Slide 45 text

Hardware tests with containers? ● It works ● You can run containers in privileged mode ● You can use drivers from host OS ● It works on a success path ● BUT: software errors, timeouts, access conflicts, device hanging when switching containers ● BUT: you still need special test equipment and drivers Expectations Reality There are limitations, but doable

Slide 46

Slide 46 text

Containers for Simulations ● We avoid using Hardware when possible ● Low overhead and startup time ● Resource quota management ○ CPU/RAM ○ Prioritization for users ○ Licenses ● Multi-instance tests (Docker Compose, Kubernetes Pods, etc.) 46

Slide 47

Slide 47 text

QEMU ● Multi-platform execution on a single machine ● Container adapters are available 47 https://github.com/qemus

Slide 48

Slide 48 text

Dev Container for Simulations 48 ● Most Simulators already have Docker images ● Docker images can be quickly converted to Dev Containers

Slide 49

Slide 49 text

Example - Verilator 49

Slide 50

Slide 50 text

And AI? 50

Slide 51

Slide 51 text

LLM Runtimes

Slide 52

Slide 52 text

LLM Runtimes Backends https://github.com/ggml-org/llama.cpp# supported-backends

Slide 53

Slide 53 text

Example - Ollama as a DevContainer Feature 53 https://github.com/prulloac/devcontainer-features/tree/main/src/ollama

Slide 54

Slide 54 text

Local Development 54 is closer to the cloud, now

Slide 55

Slide 55 text

CloudBees Office, Neuchatel, ~2015 IDEs

Slide 56

Slide 56 text

We were doing CDEs before they were a mainstream 56 Image source: https://www.combell.com/en/blog/w hat-is-rdp-remote-desktop-protocol/

Slide 57

Slide 57 text

Cloud Developer Environments ● Most of the functionality runs in the cloud ● Local Tools / IDEs are just wrappers ● SaaS, public cloud or self-hosted ● Resources and licenses can be shared 57

Slide 58

Slide 58 text

58 Embedded Projects DYI CDE suitable for your projects

Slide 59

Slide 59 text

Existing CDEs ● GitHub Codespaces ● Eclipse Che / Theia ● Ona (fka Gitpod) ● AWS Cloud9 ● Strong Network (Citrix) ● Coder ● Cloudomation ● … ● Web based AI IDEs 59 Source: https://discord.com/blog/how-discord-moved-engin eering-to-cloud-development-environments

Slide 60

Slide 60 text

Example - GitHub Codespaces 60

Slide 61

Slide 61 text

Example - GitHub Codespaces 61

Slide 62

Slide 62 text

No content

Slide 63

Slide 63 text

Developer Experience 63 May require a custom CDE

Slide 64

Slide 64 text

Existing CDEs, but for Embedded Projects 64 ● IDE Extensions ● Dev Container(s) for… ○ Build Tools ○ Simulators ○ Verification tools ● Testcontainers

Slide 65

Slide 65 text

65 Embedded Projects DYI CDE suitable for your projects

Slide 66

Slide 66 text

DIY CDE Cluster 66 Source: https://medium.com/@elliotgraebert/laptop-development-is-dead -why-remote-development-is-the-future-f92ce103fd13 Your Cloud FPGA as a service

Slide 67

Slide 67 text

DYI CDEs ● You can build a CDE on your own ● IDEs with CDE plugins ● Control planes ● Workspaces & Cloud ● Storage / Data (!) Security aspect is an area for commercial CDEs 67 Source: https://coder.com/

Slide 68

Slide 68 text

DYI CDEs ● Dev Containers and compatible IDEs ● Container Engines (e.g. Docker) and Kubernetes for the environment ● Device Plugins for FPGA (e.g. one for Xilinx) ● KubeVirt for virtual machines 68

Slide 69

Slide 69 text

llm-d for Kubernetes-native and distributed LLM inference 69 https://developers.redhat.com/articles/2025/05/2 0/llm-d-kubernetes-native-distributed-inferencing

Slide 70

Slide 70 text

Example - Xilinx Tools FPGA grids can be served and used in Kubernetes CDEs can be provisioned accordingly https://github.com/Xilinx/FP GA_as_a_Service 70

Slide 71

Slide 71 text

THE END?

Slide 72

Slide 72 text

Demos 72 ● Quickstart: ○ https://github.com/microsoft/vscode-remote-try-cpp ● My Example - Testcontainers for C/C++ ○ github.com/testcontainers/testcontainers-c ○ Testcontainers can run in Dev Containers ○ You can use GitHub Codespaces too

Slide 73

Slide 73 text

Takeaways 1. Consider using Dev Containers 2. There are tools and frameworks for embedded projects, but a lot of DYI 3. For common software stacks, the existing IDEs are at your service 4. CDEs may mean local, too

Slide 74

Slide 74 text

One First Step FOR YOU After the talk: 1. Run demos locally or in Codespaces 2. Create a Dev Container for your project 3. Share it with your team(s) 74

Slide 75

Slide 75 text

Contacts: E-mail: [email protected] GitHub: oleg-nenashev Twitter: @oleg_nenashev QUESTIONS? 75 #OpenToWork

Slide 76

Slide 76 text

Abstract 76 Back in 2017 at QA&Test Bilbao, I explored how Docker and Jenkins pipelines could transform CI/CD for embedded projects. Fast forward to 2023, containerization has become the norm in embedded software QA. My recent talk focused on tools like Testcontainers and WireMock, now widely used for testing not just backend services but entire embedded systems—from SCADA to IoT—within virtualized environments powered by Kubernetes. These technologies now form the backbone of modern testing infrastructure in software, and now expand to edge and embedded systems to cover full-system virtualization and integration testing. Same happens on the developer machines. Embedded QA Engineers are entering a new era of development enabled by Cloud Development Environments (CDEs) like GitHub Codespaces or Devcontainers, where fully configured workspaces can run on laptops, Kubernetes clusters or public clouds. Containers and VMs, often managed with tools like KubeVirt and QEMU, and wrapped by frameworks like Testcontainers, enable scalable, hardware-aware test setups that support HIL testing, firmware validation, and system modeling. It might sound like a miracle, but it’s actually possible with modern tech stacks! This talk demonstrates how to build such “hybrid developer environments” entirely with open source tools, combining C++/SystemVerilog development, embedded hardware access, and even AI-driven test automation – showing that the modern IDE for embedded QA is now distributed, intelligent, and platform-agnostic. Stay a while and listen, I have a PoC ready!