Divide & Conquer

The day job continues to keep me busy, however, I managed to connect the dots with the compute infrastructure. Input document collections now happily cascade through a chain of processes, wrapped up as lambda functions. Meta data is pushed into dynamoDB tables, storing input and output state, allowing for browsing and downloading of outputs using presign URLs eg., in node.js to produce a link that expires after two days :

var url_params = {Bucket: 'bucketname', Key: 'filename', Expires: 3600*24*2};
var url = s3.getSignedUrl('getObject', url_params);

So the vision for the backend is much closer to reality… now on to a frontend 🙂