文件列表:

Dockerfile (922, 2019-07-25)
LICENSE (1069, 2019-07-25)
az-pipeline-templates/ (0, 2019-07-25)
az-pipeline-templates/deploy-to-aks.yml (192, 2019-07-25)
az-pipeline-templates/docker-build-tag.yml (543, 2019-07-25)
azure-pipelines.yml (936, 2019-07-25)
breakdown-local.sh (128, 2019-07-25)
code/ (0, 2019-07-25)
data/ (0, 2019-07-25)
docker-entrypoint.sh (275, 2019-07-25)
environment.yml (154, 2019-07-25)
helm/ (0, 2019-07-25)
helm/mac-storage.yaml (130, 2019-07-25)
helm/vscode/ (0, 2019-07-25)
helm/vscode/.helmignore (342, 2019-07-25)
helm/vscode/Chart.yaml (102, 2019-07-25)
helm/vscode/azure-values.yaml (595, 2019-07-25)
helm/vscode/templates/ (0, 2019-07-25)
helm/vscode/templates/deployment.yaml (2121, 2019-07-25)
helm/vscode/templates/service.yaml (533, 2019-07-25)
helm/vscode/values.yaml (564, 2019-07-25)
local-deploy.sh (145, 2019-07-25)

Example template that spins up a self-contained data science environment within a container, that includes: 1. Conda for Python dependencies 2. VSCode for full featured development 3. JupyerLab for quick protyping Completely based on [Coder's](https://coder.com/) incredible work. Their GitHub repo here: https://github.com/codercom/code-server #### Why is This Useful? 1. You can develop all your code in a fully specified environment, which makes it much easier to reproduce and deploy models and analysis. 2. You can (after enabling security) move your IDE to the data. Instead of transferring data back and forth you can develop where your data is stored. 3. Last - and most important for me in industries like my own, you work with highly regulated data that has to be stored securely, where having multiple copies of data on multiple laptops can pose an unacceptably large risk. Running containers like this within a secure environment with access to the data helps us to have an ideal development environment, while ensuring the protected data remains in a secure, single location with no unnecessary duplication. #### How to Use This All Clone this down and rename the folder to be your project name. Modify the `environment.yml` file to include all the Python packages you need. Say you rename the folder to `vscode` - run the following: ``` cd ide-template docker build -t vscode . docker run -p 8443:8443 -p 8888:8888 -v $(pwd)/data:/data -v $(pwd)/code:/code --rm -it vscode ``` This will spin up the container - starting up JupyerLab and VSCode. VSCode will be running on: http://localhost:8443 JupyterLab will be running on: http://localhost:8888 with a token of `local-development` #### VSCode Extensions and Configuration You can install any extension and modify configuration like you would locally. Any extensions you install and global configuration you update will persist in the `./data` folder so you don't have to redo it every time you restart the container. By default VSCode will start up with the `./code` folder as the workspace, which you can change by modifying the `docker-entrypoint.sh` file. You can pretty much VSCode as you would locally, doing things like starting up terminals, setting Python formatters/linters, and so on. #### Deploying to kubernetes ``` docker build -t vscode . ./local-deploy ``` This will spin up the your kubernetes cluster - starting up JupyerLab and VSCode. VSCode will now be running on: http://localhost:1080 JupyterLab will be running on: http://localhost:1081 with the same token of `local-development`

近期下载者：

相关文件：

评论：[我要评论] [举报此文件]

收藏者：