Skip to content

Latest commit

 

History

History
106 lines (77 loc) · 3.37 KB

README.md

File metadata and controls

106 lines (77 loc) · 3.37 KB
Raw

Stravad

build status

Easily visualize and explore your strava activities in glorious vectorized 3D

Topo map Satellite map
Video demo
stravad-demo.webm

What is it?

Stravad periodically fetches your activities from Strava so you can work with the data locally, and easily visualize it in with the included interactive 3D map webapp. From the app, you can quickly filter activities by year, type, and so on, and click on individual activities to be taken to the strava page for them, making it easy to find old routes.

Activities are made available as:

  • GeoJSON
  • Raw strava JSON, with all the extra bits and bobs
  • A mbtiles vector tileset

Stravad is made up of two components:

  • A ruby application that uses the Strava API to download your strava data and convert it to vector tiles
  • A react frontend application, using the mapbox-gl-js library to provide interactive exploration of your data

These are packaged up for ease-of-use into docker containers, and can be run anywhere docker runs. Scripts are included to run the components directly with docker, and a Kubernetes manifest is provided for a more complete alternative.

Requirements

1. Mapbox Account

You need a Mapbox API key to use MapboxJS, and we use the Mapbox servers to retrieve our base tilesets - satellite imagery, roads, points-of-interest, and so on.

Sign up for a Mapbox account, and note down your public access key.

2. Strava Developer Account

You need API credentials to access your strava account to scrape your activity data.

From your strava account, jump to the API management page. Note down your Client ID and refresh token - you'll need them later.

Getting Started

# Make Mapbox API key available for the react build
cat > frontend/.env << EOF
REACT_APP_MAPBOX_API_TOKEN=pk.aabbccdd
EOF 

# Create a secret for the mapbox credentials
kubectl -n stravad create secret generic mapbox-credentials --from-env-file=frontend/.env

# Fetch a refresh token for strava. You'll need to plug
# the response in further down!
export STRAVA_CLIENT_ID=123
export STRAVA_CLIENT_SECRET=00aabbccddee
cd fetch
./fetch-token-with-docker.sh
cd ..

# Create an `env.sh` in `fetch` to specify strava access credentials:
cat > fetch/env.sh << EOF
STRAVA_CLIENT_ID=$STRAVA_CLIENT_ID
STRAVA_CLIENT_SECRET=$STRAVA_CLIENT_SECRET
STRAVA_REFRESH_TOKEN=eeffaabbccdd
EOF

# Create the namespace
kubectl create namespace stravad

# Create a secret containing the environment variables we need
kubectl -n stravad create secret generic strava-credentials --from-env-file=fetch/env.sh

# Apply the manifest
kubectl apply -f deploy/manifest.yaml

Once everything's up and running, we can manually trigger the scraping job, rather than waiting until the daily cron trigger fires:

kubectl create job --from=cronjob/stravad-scrape runnow -n stravad