The Array of Things File Browser (AFB) is a multi-step process and application. It consists of a handful of data preprocessors that run scripts on a server within the AoT realm and the web application lives out on AWS.
The web app is a standard Phoenix application. There's a handful of routes that display information and a single route responsible for getting data from S3 to update the database's cache of data set metadata.
There is a single Python script (taken from the Waggle data tools) and a series of bash scripts to coordinate processing batches.
The processes are kicked off by cron. There are daily, weekly, biweekly, and
monthly jobs. Each of these are coordinated by the script named
cron-${epoch}.bash that in turn call the slicing bash script. These scripts
download tarballs, call the Python slicing script, repackage the slices and
upload them to S3.
The daily script also calls the processing route of the web application to kick off the metadata refresh routine.