First Login¶
After installing AKKO, this guide walks you through logging in to the main services and understanding the SSO flow.
The Cockpit¶
Open the AKKO cockpit -- the central portal for all services:
The cockpit displays a card for every running service with real-time health status, latency, and direct links. It also provides a Quick Start bar and a Tech Stack overview showing exact versions of every component.
Bookmark the cockpit
The cockpit is your single entry point. From here, you can click through to any service without remembering individual URLs.
SSO with Keycloak¶
AKKO uses Keycloak as its identity provider. All services share a single sign-on (SSO) realm called akko.
How it works:
- Click any service link in the cockpit (e.g., Superset, Airflow, Dashboards)
- You are redirected to the Keycloak login page
- Enter your username and password
- After authentication, you are redirected back to the service
- Subsequent services recognize your session -- no need to log in again
Default Users and Roles¶
AKKO comes with four pre-configured test users, each with a different RBAC role:
| User | Password | Role | Access Level |
|---|---|---|---|
alice |
alice123 |
akko-admin |
Full access to all services and data |
bob |
bob123 |
akko-engineer |
Read/write access to data |
carol |
carol123 |
akko-analyst |
Read-only access to data |
dave |
dave123 |
akko-viewer |
Minimal access (dashboards only) |
Use alice for getting started
For your first exploration, log in as alice (alice123). She has admin
privileges on all services, so you will not hit any permission issues while
learning the platform.
Logging In to JupyterHub¶
- Navigate to https://lab.akko.local
- Click Sign in with Keycloak (or you will be redirected automatically)
- Enter
alice/alice123 - JupyterHub spawns a personal notebook container with:
- JupyterLab -- Python, R, and Julia kernels
- code-server (VS Code in the browser)
- Quarto for publishing reports
- Pre-loaded notebooks in the
notebooks/directory (read-only)
Kernel selection
After JupyterHub spawns your server, select the Python 3 kernel for the banking demo and most notebooks. The R and Julia kernels are available for the analytics notebooks.
Logging In to Superset¶
- Navigate to https://bi.akko.local
- Log in with
alice/alice123via Keycloak SSO - A dashboard and datasets are auto-provisioned by the
superset-initsidecar:- 8 datasets connected to Trino (Iceberg tables)
- 8 charts (KPIs, bar charts, time series, pie charts)
- 1 dashboard: AKKO Banking Overview
Dashboard requires data
The auto-provisioned dashboard will show empty charts until you run the banking demo notebook, which creates and populates the Iceberg tables that Superset queries via Trino.
To explore the dashboard:
- Click Dashboards in the top navigation
- Open AKKO Banking Overview
- After running the banking notebook, refresh the dashboard to see live data
Logging In to Airflow¶
- Navigate to https://orchestrator.akko.local
- Log in with
alice/alice123via Keycloak SSO - You will see the E2E pipeline DAG (
akko_e2e_pipeline):- Demonstrates the full data flow: ingest, transform, quality checks
- Uses the OpenLineage provider for lineage tracking
Triggering the DAG
The E2E DAG is paused by default. Toggle it on and click Trigger DAG to run it manually. Check the task logs to see each step in action.
Logging In to Dashboards¶
- Navigate to https://metrics.akko.local
- Log in with
alice/alice123via Keycloak SSO - Pre-configured dashboards monitor the platform:
- Container metrics (CPU, memory, network)
- Trino query performance
- Service health overview
Dashboards also provides logs layer integration for centralized log viewing. Navigate to Explore and select the logs layer data source to search container logs across all services.
Logging In to object storage¶
- Navigate to https://storage.akko.local
- Log in is protected by oauth2-proxy (SSO via Keycloak)
- Browse the
akko-warehousebucket -- this is where all Iceberg table data is stored
Keycloak Admin Console¶
To manage users, roles, and SSO clients:
- Navigate to https://identity.akko.local
- Click Administration Console
- Log in with:
- Username:
admin - Password: check the Keycloak admin password in your Helm values or Kubernetes Secret
- Username:
Do not confuse admin accounts
The Keycloak admin account (admin) is separate from the AKKO test users
(alice, bob, carol, dave). The Keycloak admin manages the SSO realm itself,
while test users access the analytics stack services.
From the Keycloak admin console, you can:
- View and edit the akko realm
- Manage users (add, disable, reset passwords)
- Configure SSO clients (13 pre-configured clients for all AKKO services)
- View active sessions and audit login events
Other Services¶
| Service | URL | Notes |
|---|---|---|
| Trino UI | https://federation.akko.local |
Query monitoring (protected by SSO) |
| Spark Master | https://compute.akko.local |
Spark cluster overview |
| Polaris | https://polaris.akko.local |
Iceberg catalog management API |
| Prometheus | https://prometheus.akko.local |
Metrics and alerting rules |
| Traefik | https://traefik.akko.local |
Reverse proxy dashboard |
| AKKO Docs | https://docs.akko.local |
Documentation and published reports |
| OpenMetadata | https://catalog.akko.local |
Data catalog (governance profile) |
Next Steps¶
Now that you are logged in, proceed to the First Notebook guide to create your first Iceberg tables and run a federated query across the platform.