Skip to main content
The iii Console is a standalone application that gives you full operational visibility into a running iii engine. It provides a web UI to inspect functions, triggers, state, streams, traces, and logs — all in real time.

Quick Start

The console comes preinstalled with iii, start it with:
Then open your browser to http://127.0.0.1:3113.
The console connects to a running iii engine instance. Make sure your engine is started before launching the console. By default it expects the engine at 127.0.0.1:3111.
The root URL redirects to the Workers page. The sidebar includes Workers, Functions, Triggers, States, Streams, Queues, Traces, Logs, and Config. The optional Flow page appears when flow visualization is enabled.

Workers

The Workers page lists every worker process currently connected to the engine, with live metrics (memory, CPU, event-loop lag), connection time, active invocations, and registered functions. iii Console Workers page showing connected workers grouped by runtime with status, PID, IP, and active invocation counts The worker detail panel shows full metadata, telemetry (language, project, framework), live metrics, and the list of functions the worker has registered. iii Console Workers page with a selected worker detail panel showing runtime, version, OS, metrics, and registered functions
Workers launched via iii worker add ./path or iii worker add <image> run inside libkrun microVMs. The libkrun launcher builds the VM boot environment and injects III_ISOLATION=libkrun at VM boot time, so you’ll see a purple LIBKRUN badge next to the runtime badge. The badge is descriptive (what the worker reports about its environment), not an attestation — a direct SDK connection can self-set III_ISOLATION=docker (or anything else) before starting. Workers that don’t set the variable show no isolation badge.

Functions

The Functions page lists every function registered with the iii engine and lets you invoke any of them directly with custom JSON input. iii Console Functions page showing grouped functions and the invoke panel for a selected function All registered functions are displayed with their metadata: Use the search bar at the top to filter functions by name.

Invoking Functions

The detail panel includes a JSON editor for invoking functions directly. Enter a payload, click Invoke, and view the result inline. The console sends requests via POST /_console/invoke.
See Use the Console for step-by-step instructions.
Only functions that are currently registered and connected via an active worker will appear in the list. If a function is missing, check that its worker process is running.

Triggers

The Triggers page shows every trigger registered with the engine and provides interactive tools to test each trigger type. Triggers are grouped by kind (http, cron, event, state) with a total count badge and per-type filter tabs. iii Console Triggers page showing HTTP triggers and the request builder for testing a selected endpoint

Testing Triggers

The console provides interactive testing tools for each trigger type:
See Use the Console for step-by-step instructions on testing each trigger type.
The available trigger types depend on which workers are loaded in your engine configuration. See Trigger Types and Workers for the full list.

States

The States page provides a browser for the engine’s key-value state store. You can view, create, edit, and delete state entries organized by scope. The layout is divided into three panels: a group list on the left, an items table in the center, and a detail sidebar on the right. iii Console States page showing state groups, key-value items, and JSON value details for a selected item The state browser displays a two-level hierarchy:
  1. Groups (Scopes) — top-level namespaces that organize state
  2. Items — individual key-value pairs within each group
Select a group from the left panel to see all its key-value items. Each item shows a Key and Value (rendered as formatted JSON). Complex values are displayed in a collapsible JSON viewer.

Managing State

The console supports full CRUD operations on state entries:
See Use the Console for step-by-step instructions.
State persistence depends on your engine’s State worker configuration. With in_memory storage, state is lost on engine restart. With file_based or RedisAdapter, state persists across restarts. See State Worker for configuration details.
Use the search bar to filter items by key name. For groups with many items, the browser supports pagination.

Streams

The Streams page is a live WebSocket monitor that captures messages flowing through the engine’s stream connections. It shows message counters (total, inbound, outbound, and buffer size), subscription management, and direction filter tabs. Use the filter toggle to show or hide system streams. Streams update in real time via WebSocket.
Streams are provided by the Stream worker. Make sure iii-stream is included in your engine configuration. See Stream Worker for details.

Queues

The Queues page lists durable queue topics, subscriber counts, queued message counts, in-flight work, retry counts, and dead-letter activity. iii Console Queues page showing durable topics, subscriber counts, dead-letter counts, and the selected topic overview panel The topic detail panel has two tabs:
See Use the Console for step-by-step instructions on testing queues and managing dead letters.

Traces

The Traces page provides full OpenTelemetry trace visualization with multiple view modes, advanced filtering, and detailed span inspection. iii Console Traces page showing a waterfall chart for a selected trace and span timing details
Trace collection requires the Observability worker with exporter set to memory or both. See Observability Worker for configuration.

Trace List

View Modes

The trace detail view provides five visualization modes: iii Console Traces page showing the flame graph view for a selected trace iii Console Traces page showing a service trace map with connected services for a selected trace iii Console Traces page showing the flow graph view for a selected trace

Span Details

The span detail panel includes the following tabs:

Trace Filtering

Multiple filters are combined with AND logic. Pagination controls at the bottom allow browsing large result sets.
See Use the Console for step-by-step instructions on inspecting traces and spans.

Logs

The Logs page provides a viewer for structured OpenTelemetry logs collected by the engine. Logs are displayed in reverse chronological order. Each entry shows a timestamp, severity level, service name, trace/span context, and message. A severity filter toggle, full-text search, and time-range controls let you zero in on specific log entries. The source breakdown at the bottom shows which services are contributing the most log volume. iii Console Logs page showing structured log rows and a selected log detail panel with context data
Log collection requires the Observability worker with logs_enabled: true. See Observability Worker for configuration.
The iii Console (this web UI) is separate from the engine’s logs_console_output setting. The logs_console_output option controls whether OTLP logs are also printed to the engine process’s stdout — it does not affect the Logs page here.

Log Viewer

Expanding a log entry reveals the full JSON payload, including attributes, trace/span IDs, and resource metadata.

Log Filtering

Log Entry Details

Each expanded log entry includes:
If a log entry has a trace_id, you can click it to jump directly to the corresponding trace in the Traces section above.

Flow

The Flow page renders an interactive graph of your system’s architecture, showing how triggers, functions, state stores, and queues connect.
The Flow page is an opt-in feature. Enable it by starting the console with the --enable-flow flag or setting the III_ENABLE_FLOW environment variable.
The flow diagram uses an auto-layout algorithm (Dagre) to arrange nodes and edges: Edges show the data flow direction between components — from triggers to the functions they invoke, and from functions to the state or queues they interact with. The graph uses auto-layout (Dagre) and supports pan, zoom, and node inspection. Layout configuration is saved to the engine’s state store and restored on next visit.
See Use the Console for navigation instructions.

Configuration

The iii Console is configured via CLI flags and environment variables. All settings have sensible defaults for local development. iii Console Config page showing engine endpoints, trigger types, active workers, trigger handlers, and worker pools

CLI Flags

Environment Variables

Default Ports

Examples

The console includes a Config page (accessible from the sidebar) that displays the current runtime configuration: engine connectivity, port information, OpenTelemetry settings, and service version.