Skip to content
Chatto v0.5.0

Read This First

Chatto ships as one compact binary. The chatto run command starts a Chatto server.

This guide helps you select the standalone binary, Docker Compose, or Kubernetes.

ComponentRequired forWhat to protect
Chatto serverWeb client, API, and background workConfiguration and secrets
NATS with JetStreamDurable server data; embedded or externalPersistent storage and encryption keys
S3-compatible storageOptional external file storageUploaded objects and storage credentials
Search providerOptional full-text searchIndex with text from decrypted messages
Asset processing workerOptional video and GIF conversionCPU capacity and temporary disk space
LiveKitOptional voice and video callsCredentials, public media ports, and TLS

Search and asset processing can run inside the Chatto server or as separate processes. Clients send call media directly to LiveKit. See the configuration guides for each optional service.

Chatto does not use MySQL or PostgreSQL. Chatto uses NATS and JetStream to store durable event data and runtime data.

Chatto can start an embedded NATS server. Chatto can also connect to an external NATS server.

Use external NATS when you run more than one Chatto server process. The external service lets all Chatto server processes use the same durable data. Use external NATS when high availability or updates without downtime are necessary.

Embedded NATS stores its data in the local filesystem. Use it only for one Chatto server process. Your server will stop briefly when you update Chatto. Use the Chatto backup and restore commands to migrate between these configurations.

Chatto stores uploaded files in NATS Object Store by default. Use this configuration for a small server that stores a small quantity of uploaded data.

For most production servers, use an S3-compatible object store. The object store can be a hosted service or a service that you operate. It must be compatible with the AWS S3 SDK.

Read Media & Attachments for the storage and processing configurations.

Each Chatto server has separate user accounts and community data. Users can register directly or use an external identity provider. An account on one Chatto server is not an account on a different server.

Chatto encrypts message text and selected account fields before it writes them to durable storage. Chatto does not encrypt each record or uploaded asset.

Read Encryption at Rest and Data Erasure for the full protection boundary. Chatto includes a built-in key management service (KMS). You cannot configure an external KMS currently.

Chatto includes calls and screen sharing through LiveKit. Deploy LiveKit with Chatto to enable these features. Each deployment guide gives the necessary configuration.

  • Public URL: Select a domain, such as chat.example.com, that your users can access.
  • TLS: Use HTTPS for an Internet-facing production server. Prepare a TLS certificate or use a deployment that gets one automatically.
  • NATS: Use embedded NATS for one Chatto server process. Use external NATS for multiple processes.
  • Data storage: Put the NATS data directory on durable storage. Monitor the available disk space.
  • File storage: Use NATS Object Store when users upload a small quantity of data. Use S3-compatible storage when users upload more data.
  • Owner account: Add each owner email address to owners.emails. Chatto gives the owner role after a user verifies a matching address. This setting does not create an account.
  • Login methods: Select direct registration, external login providers, or the two methods. Do a test of an external provider before you set auth.direct_login to false.
  • Transactional email: Configure SMTP (the default transport) or JMAP for direct registration, email verification, and other email features. Do a test of the credentials before you invite users.
  • Backups: Select storage for data backups, key exports, and S3 assets. Do a test of the restore procedure at regular intervals.