Inngest Factories

Deploy Inngest on Kubernetes using the official Helm chart. Inngest is a workflow orchestration platform with no CRDs — all configuration is through Helm values.

Import

import { inngestBootstrap } from 'typekro/inngest';

Quick Example

import { inngestBootstrap } from 'typekro/inngest';

// 'kro' = KRO mode (continuous reconciliation via ResourceGraphDefinition)
// 'direct' = Direct mode (immediate apply, no KRO controller needed)
const factory = inngestBootstrap.factory('kro', {
  namespace: 'inngest',
  waitForReady: true,
});

await factory.deploy({
  name: 'inngest',
  namespace: 'inngest',
  inngest: {
    eventKey: 'your-event-key',
    signingKey: 'your-signing-key',
  },
});

Available Factories

Factory	Kind	Description
inngestHelmRepository	HelmRepository	OCI chart registry
inngestHelmRelease	HelmRelease	Inngest deployment via Helm

Bootstrap Composition

Basic (bundled PostgreSQL + Redis)

await factory.deploy({
  name: 'inngest',
  namespace: 'inngest',
  inngest: {
    eventKey: 'your-event-key',
    signingKey: 'your-signing-key',
  },
});

With external databases (CNPG + Valkey)

Disable bundled databases and provide connection URIs:

await factory.deploy({
  name: 'inngest',
  namespace: 'inngest',
  inngest: {
    eventKey: 'your-event-key',
    signingKey: 'your-signing-key',
    postgres: { uri: 'postgresql://inngest:password@my-db-rw:5432/inngest' },
    redis: { uri: 'redis://my-cache:6379' },
    sdkUrl: ['http://my-app:5173/api/inngest'],
  },
  postgresql: { enabled: false },
  redis: { enabled: false },
  resources: {
    requests: { cpu: '500m', memory: '1Gi' },
    limits: { cpu: '2', memory: '4Gi' },
  },
});

With ingress and autoscaling

await factory.deploy({
  name: 'inngest',
  namespace: 'inngest',
  inngest: {
    eventKey: 'your-event-key',
    signingKey: 'your-signing-key',
    host: 'inngest.example.com',
  },
  ingress: {
    enabled: true,
    className: 'nginx',
    hosts: [{ host: 'inngest.example.com' }],
    tls: [{ secretName: 'inngest-tls', hosts: ['inngest.example.com'] }],
  },
  keda: {
    enabled: true,
    minReplicas: 2,
    maxReplicas: 20,
  },
});

Bootstrap Status

instance.status.ready    // boolean — Inngest is processing events
instance.status.phase    // 'Ready' | 'Installing'
instance.status.failed   // boolean — true if HelmRelease Ready=False
instance.status.version  // deployed chart version (static, deploy-time)

Note: phase cannot distinguish 'Failed' from 'Installing' due to a CEL evaluator limitation. Use the failed field to detect deployment failures.

Key Configuration

Field	Required	Description
inngest.eventKey	Yes	Event authentication key
inngest.signingKey	Yes	Request signing key
inngest.postgres.uri	No	External PostgreSQL URI (disables bundled PG)
inngest.redis.uri	No	External Redis/Valkey URI (disables bundled Redis)
inngest.sdkUrl	No	SDK URLs to auto-sync functions from
postgresql.enabled	No	Bundled PostgreSQL (default: true)
redis.enabled	No	Bundled Redis (default: true)

Usage in Compositions

Note: This example uses CloudNativePG and Valkey for external databases. See their respective docs for setup.

import { type } from 'arktype';
import { kubernetesComposition } from 'typekro';
import { inngestBootstrap } from 'typekro/inngest';
import { cluster } from 'typekro/cnpg';
import { valkey } from 'typekro/valkey';

const MyPlatform = kubernetesComposition({
  name: 'my-platform',
  kind: 'MyPlatform',
  spec: type({ name: 'string', image: 'string' }),
  status: type({ ready: 'boolean' }),
}, (spec) => {
  // PostgreSQL via CNPG
  const db = cluster({
    id: 'database',
    name: `${spec.name}-db`,
    spec: {
      instances: 3,
      storage: { size: '50Gi' },
      bootstrap: { initdb: { database: 'inngest', owner: 'inngest' } },
    },
  });

  // Valkey cache
  const cache = valkey({
    id: 'cache',
    name: `${spec.name}-cache`,
    spec: { shards: 3, volumePermissions: true },
  });

  // Inngest with external DBs
  const inngest = inngestBootstrap({
    name: `${spec.name}-inngest`,
    inngest: {
      eventKey: 'your-key',
      signingKey: 'your-signing-key',
      postgres: { uri: `postgresql://inngest:password@${spec.name}-db-rw:5432/inngest` },
      redis: { uri: `redis://${spec.name}-cache:6379` },
    },
    postgresql: { enabled: false },
    redis: { enabled: false },
  });

  return {
    ready: db.status.readyInstances > 0 && cache.status.ready,
  };
});

Prerequisites

No operator installation needed — Inngest is deployed directly via Helm. Requirements:

• Flux CD — for HelmRelease reconciliation
• PostgreSQL — bundled or external (e.g., CloudNativePG)
• Redis/Valkey — bundled or external (e.g., Valkey)
• KEDA — optional, for queue-depth autoscaling
• cert-manager — optional, for TLS ingress

Next Steps

• CloudNativePG — External PostgreSQL clusters
• Valkey — External Valkey cache clusters
• Kubernetes Factories — Core Kubernetes resources
• Inngest Self-Hosting Docs — Upstream reference