Skip to content

Latest commit

 

History

History
176 lines (118 loc) · 5.82 KB

getstarted.md

File metadata and controls

176 lines (118 loc) · 5.82 KB
Raw

Quick up-and-running

The aim is to establish various demos of software solutions for the MyPCH project around the Tidepool platform (TP) (https://www.tidepool.org) and Semantic Container (SC) (https://www.ownyourdata.eu/en/semcon/).

Prerequisites

sudo apt-get update
sudo apt-get install docker-ce docker-ce-cli containerd.io

Add your user to the docker group, so docker commands can be run without sudo:

sudo usermod -aG docker $USER
sudo newgrp docker  # will activate this for your user, without a reboot

Tidepool

From Tidepool we use the development repo to get the platform at a local instans. The uploader repo contains the an extract utility

Steps for creating repos

  1. Create a new folder, e.g tidepool-development, to hold all the Tidepool repos.
mkdir tidepool-development
cd tidepool-development
  1. Clone Tidepool development, uploader, blip

Recommend to use the same names as Tidepool to avoid changing in code.

git clone https://github.com/tidepool-org/development development
git clone https://github.com/tidepool-org/uploader uploader
git clone https://github.com/tidepool-org/blip blip
  1. Set path
    Avoid PATH problems. Use your ~/.bashrc file. Remember to be in under tidepool-development repository
cd tidepool-development
echo "export PATH=${PWD}/bin:${PATH}" >> ~/.bashrc
exec bash

Steps to get it up running

The process to get Tidepool running with a local Kubernetes cluster is well described https://github.com/tidepool-org/development/tree/k8s-stable Please remember that you need to check out this branch.

git fetch -ap
git checkout -b k8s-stable origin/k8s-stable

Different versions were listed in the documentation of Tidepool. This is just copying it here:

# Linux
curl -LO https://storage.googleapis.com/kubernetes-release/release/v1.15.1/bin/linux/amd64/kubectl
chmod +x ./kubectl
sudo mv ./kubectl /usr/local/bin/kubectl

# Linux
curl -fsSL https://get.helm.sh/helm-v3.0.2-linux-amd64.tar.gz | tar -xzv linux-amd64 && sudo mv linux-amd64/helm /usr/local/bin/helm

# Linux
curl -fsSL https://github.com/windmilleng/tilt/releases/download/v0.11.4/tilt.0.11.4.linux.x86_64.tar.gz | tar -xzv tilt && sudo mv tilt /usr/local/bin/tilt

You need also to create a directory for MongoDB data.

export TIDEPOOL_DOCKER_MONGO_VOLUME="~/MyMongoData"

The process with the tidepool helper scripts involves:

tidepool server-init

Please remember to add export KUBECONFIG=${HOME}/.kube/config to your .bashrc file, and do exec bash to reload the configuration. You might need to do mkdir ~/.kube first.

Smoketest web

tidepool start will bring everything up, and you can access http://localhost:3000 and then create a user.

Create user account

The account verification can be done using:

tidepool verify-account-email <email>

Debug

The tidepool services can behave a bit oddly. In this case, it may be required to delete the images. The tidepool utility provides a wrapper for docker to prune/delete the relevant images. The images and containers are stateless.

Check prerequisites:

tidepool doctor

Stop server

tidepool stop (in a new terminal)
tidepool server-prune (if stop not works)

Start server:

tidepool server-init
tidepool start

Uploader

The uploader service needs to be prepared.

The Tidepool uploader can easy be launched on a Linux platform using the Docker container, thanks to Tim (@TheDukeDK) and Mads (@atombrella) from Denmark.

If you are running Linux you probably need to be using an Ubuntu distribution or derivative. To get around this for other distrubutions you can try to build a local docker image which is based on Ubuntu 18.04 and use the yarn/npm commands interactively.

The process is described in the documentation https://github.com/tidepool-org/uploader#docker-for-linux

NOTE: You need to add udev rules to your host for uploads to actually work. You can find the udev rules here. The file should be placed in /etc/udev/rules.d/ and the host should be rebooted.

  1. Build the image docker-compose build

  2. Run it docker-compose up -d

Making a new Windows or MacOS package is feasible with

docker exec -it uploader bash -c "yarn package-win"  # or package-mac or package-all

It should leave the files in the host.

Smoketest uploader

Test on the client and work with it interactively.

Even if you kill the Tidepool Uploader GUI the container will continue to run. You can work with the yarn commands like you would locally by using docker exec.

Examples

Interactively select the yarn target: docker exec -it uploader bash -c "yarn run"

Rebuild: docker exec -it uploader bash -c "yarn build"

Start the Dev GUI: docker exec -it uploader bash -c "yarn dev"

Semantic Container

The fork in MyPCH https://github.com/MyPCH/sc-diabetes is based on the original upstream source that can be found at https://github.com/sem-con/sc-diabetes

Clone the sc-diabetes semantic container repo from orginal upstream to get the newest version:

git clone https://github.com/sem-con/sc-diabetes.git

SC demonstrations

In the folder dataflows the demonstation of SC capabilities can found at these links: