SlimFaas User Interface

SlimFaas includes a web user interface that gives operators a live view of functions, jobs, queues, and network activity.

The user interface is available at the SlimFaas root address, for example:

http://<slimfaas>/

The UI is served by the SlimFaas application itself and uses the same backend endpoints as the API. It is intended for operational visibility and manual wake-up actions, not as a replacement for Kubernetes configuration.

---

1. What the Page Shows

The page is split into two main sections:

• Infrastructure Overview: shows all detected SlimFaas functions and their runtime state.
• Jobs Overview: shows configured jobs, running jobs, and scheduled executions.

When the live front features are enabled, the page also displays a network map that visualizes traffic between external callers, SlimFaas, queues, functions, and SlimFaas replicas.

---

2. Infrastructure Overview

The infrastructure section lists every deployment, statefulset, daemonset, or websocket function detected by SlimFaas.

For each function, the table shows:

• Name: function name and the sync/async route pattern.
• Visibility: public/private access and trust level.
• Scale: ready/requested replicas, minimum replicas, start replicas, scale-down timeout, parallel request limits, schedules, dependencies, and autoscaling triggers.
• Resources: CPU and memory requests/limits when available.
• Replicas: pod-level status, readiness, and pod IPs.

If a function has no ready replica, the UI shows it as down and allows it to be manually woken up.

---

3. Wake-Up Actions

The UI exposes two wake-up actions:

• Wake Up on one function calls POST /wake-function/{functionName}.
• Wake Up All Functions calls POST /wake-functions.

SlimFaas coalesces repeated wake-up calls while a wake-up is already in progress for the same function. The UI also applies a short local cooldown to avoid accidental repeated clicks.

---

4. Live Status Stream

The UI connects to:

GET /status-functions-stream

This endpoint uses Server-Sent Events (SSE). It sends:

• state events: periodic full snapshots containing functions, queue lengths, jobs, SlimFaas replicas, SlimFaas nodes, and front status.
• activity events: single live network activity events.
• activity_batch events: grouped live network activity events during bursts.

The browser reconnects automatically if the stream disconnects.

---

5. Network Map

When SlimFaas:EnableFront is enabled, the UI renders a live network map.

The map uses the activity stream to show messages moving between:

• external callers
• SlimFaas
• function pods
• async queues
• SlimFaas replicas in multi-node deployments

The map is live-only for animations. Historical activity is not replayed into the animation stream when a new browser session starts.

SlimFaas nodes also synchronize recent local activity through the internal endpoint:

GET /internal/activity-events?since=<unix-ms>

This endpoint is intended for peer SlimFaas nodes inside the namespace.

---

6. Jobs Overview

The jobs section uses the same SSE state payload as the function dashboard.

It shows:

• number of job configurations
• number of currently running jobs
• number of configured schedules
• image, visibility, dependencies, resources, schedules, and running job details when available

If no job configuration is loaded, the section displays an empty state.

---

7. Main Backend Endpoints Used by the UI

---

8. Configuration in appsettings.json

SlimFaas configuration is read from the SlimFaas section in appsettings.json.

The same values can be overridden with environment variables. For .NET configuration, use __ to represent nested keys. For example:

SlimFaas__EnableFront=false
SlimFaas__StatusStream__StateIntervalMilliseconds=2000

SlimFaas UI and Dashboard Settings

StatusStream is optional in appsettings.json. If the section is missing, SlimFaas uses the default values above.

9. Example Kubernetes Environment Overrides

env:
  - name: SlimFaas__EnableFront
    value: "true"
  - name: SlimFaas__StatusStream__StateIntervalMilliseconds
    value: "1000"
  - name: SlimFaas__StatusStream__MaxSseClients
    value: "100"
  - name: SlimFaas__StatusStream__MaxLiveEventsPerSecond
    value: "500"
  - name: SlimFaas__StatusStream__LiveActivityBatchSize
    value: "100"

Use lower intervals for more reactive dashboards and higher intervals for lower backend load. In high-traffic clusters, prefer setting MaxLiveEventsPerSecond, LiveEventSamplingRatio, and LiveActivityBatchSize instead of disabling the front entirely.