Apache Airavata Portals

The airavata-portals repository is a consolidated home for all web-based user interfaces built on top of the Apache Airavata middleware platform. This collection of frontend components and frameworks enables seamless interaction with Airavata's powerful orchestration, identity, data, and compute services.

Running locally with Tilt

The Django portal runs as a container tenant on the shared airavata-devstack substrate (one colima VM, one Traefik ingress serving *.airavata.host), managed with Tilt. The two stacks run as separate Tilt instances.

One-time setup

# From either repo — both carry the identical devstack kit
./devstack/devstack setup

This installs colima, mkcert, dnsmasq, creates the shared VM and Traefik ingress, and configures wildcard DNS for *.airavata.host → 127.0.0.1 (trusted cert, no -k).

Daily startup

# 1. Start the Airavata backend stack (in the apache/airavata repo)
cd ../airavata && tilt up

# 2. Start the portals (this repo) on a distinct Tilt port
tilt up --port 10351

The Django portal is then served at https://gateway.airavata.host (trusted HTTPS via the shared Traefik ingress). The portal runs inside the shared colima VM as a container; settings_local.py is generated automatically on first tilt up if the file does not exist. Only the Django portal is wired into the Tiltfile today; other portals can be added as additional resources later.

Log in and run your first experiment (Echo)

The backend stack seeds a ready-to-use tenant outside the JVM when its database is first created — directly from SQL, before the server starts (default gateway, SFTP storage, a docker-SLURM compute resource with a normal queue, the Echo application, and a Default Project, all shared with default-admin). So once both stacks are green you can run an experiment with zero manual setup:

1. Open https://gateway.airavata.host and log in as default-admin / ade4#21242ftfd.
2. Create an experiment: choose the Echo application, the Default Project, and the slurm compute resource (queue normal). The Input_to_Echo field defaults to Hello, Airavata!.
3. Launch it — it runs on the docker-SLURM cluster (env setup → sbatch over SSH → sacct monitoring → SFTP output staging) and reaches COMPLETED, with Echo.stdout holding the echoed input.

tilt down then tilt up brings the same working state back up (the database volume persists); ./devstack/devstack reset (or wiping the db_data volume) re-seeds it from scratch.

Repository Structure

This repository contains the following sub-projects and templates:

Portals and SDKs

• airavata-django-portal
  The reference web-based user interface for interacting with Airavata services, supporting job submissions, project management, and monitoring. Talks to Airavata over gRPC via the airavata-python-sdk.

Starter Templates

• airavata-cookiecutter-django-app
  Cookiecutter template to scaffold new Django apps for integration with the Django portal.

• airavata-cookiecutter-django-output-view
  Template for building reusable output viewers compatible with portal job results.

Legacy and Other Frontends

• airavata-php-gateway
  Legacy PHP-based science gateway frontend (archived/deprecated).

• airavata-custos-portal
  Web-based UI for managing Custos identity, group, and resource permissions.

Purpose

The goal of this consolidation is to:

• Simplify the discovery and contribution process for Airavata frontend components.
• Encourage reuse of UI components through a shared ecosystem.
• Promote rapid prototyping and customization of science gateways.
• Align documentation and tooling across related UI projects.