Skip to content

Latest commit

 

History

History
77 lines (61 loc) · 3.57 KB

README.md

File metadata and controls

77 lines (61 loc) · 3.57 KB
Raw

Jupyterhub

This project will give us a container running jupyterhub with WrapSpawner which runs a MesosSpawner and DockerSpawner

Please follow the steps for Running it like in "prod" to make sure all dependencies will resolve and your changes will actually work

Run it "locally" for faster iterations

  1. Create a .env file with the content below
JPY_COOKIE_SECRET=<look for it in your secrets storage solution>
CONFIGPROXY_AUTH_TOKEN=<look for it in your secrets storage solution>
JUPYTERHUB_CRYPT_KEY=<look for it in your secrets storage solution>
OAUTH_CLIENT_ID=<look for it in your secrets storage solution>
OAUTH_CLIENT_SECRET=<look for it in your secrets storage solution>
POSTGRES_PASSWORD=<whatever you want since it is running locally>
DB_PASS=<whatever you want since it is running locally>
# Start of marathon in appdev config
HUB_CONNECT_URL=http://<IP of your computer on appdev VPN>:<port you setup>
MARATHON_HOST=http://master.mesos.service.dev.ivy:8080/
# End of marathon in dev config
LOG_LEVEL=DEBUG
DB_HOST=localhost:5432
DB_USER=jupyterhub
DB_NAME=jupyterhub
SERVICE_NAME=jupyterhub
PUBLIC_URL=http://jupyterhub.local.nxtlytics.dev:8000
PATH_TO_GOOGLE_SERVICE_ACCOUNT_JSON=/path/in/your/computer/to/service/account/json/file.json
# Start of remote docker spawner config
DOCKER_HOST=tcp://<IP of docker host>:2376
PATH_TO_DOCKER_CA=/path/in/your/computer/to/docker-ca.pem
PATH_TO_JUPYTERHUB_CLIENT_KEY=/path/in/your/computer/to/jupyterhub-local-key.pem
PATH_TO_JUPYTERHUB_CLIENT_CERTIFICATE=/path/in/your/computer/to/jupyterhub-local-cert.pem
HUB_IP=<IP of your computer in that spawned container can reach you at>
# End of remote docker config

Note: DOCKER_HOST must include tcp:// due to these lines of code

  1. Install configurable-http-proxy and set your path to find it
$ npm install configurable-http-proxy@4.2.0
$ export PATH="${PATH}:${PWD}/node_modules/.bin"
  1. Start the database first by running docker-compose up -d db
  2. cd app
  3. pipenv run jupyterhub or pipenv run jupyterhub upgrade-db && pipenv run jupyterhub if this is the first time you connect to the database
  • Note: Remember to have your hostname in /etc/hosts (Example: 127.0.0.1 jupyterhub.local.nxtlytics.dev)

Run it like in "prod"

  1. Create a secrets.env file at the root of the repository with the content below

NOTE: POSTGRES_PASSWORD and DB_PASS should have the same values, example of these can be found in 1Password in the Infrastructure vault

JPY_COOKIE_SECRET=<look for it in your secrets storage solution>
CONFIGPROXY_AUTH_TOKEN=<look for it in your secrets storage solution>
JUPYTERHUB_CRYPT_KEY=<look for it in your secrets storage solution>
OAUTH_CLIENT_ID=<look for it in your secrets storage solution>
OAUTH_CLIENT_SECRET=<look for it in your secrets storage solution>
POSTGRES_PASSWORD=<whatever you want since it is running locally>
DB_PASS=<whatever you want since it is running locally>
HUB_CONNECT_URL=http://<IP of your computer on VPN>:<port you setup>
GOOGLE_SERVICE_ACCOUNT_JSON=<content of /path/in/your/computer/to/service/account/json/file.json in one line>
DOCKER_CA=<Replace new lines with literal \n>
JUPYTERHUB_CLIENT_CERTIFICATE=<Replace new lines with literal \n>
JUPYTERHUB_CLIENT_KEY=<Replace new lines with literal \n>

NOTE: cat /path/in/your/computer/to/service/account/json/file.json | jq -c | sed 's!\\n!\\\\n!g' | pbcopy will give you the value for GOOGLE_SERVICE_ACCOUNT_JSON

  1. Run docker-compose up