pre-1.0
Documentation

How to run it

ProxDock is pre-1.0 and so are these pages. The guides below are being written as the features settle; where a page is still an outline it says so at the top rather than pretending otherwise.

Guides

Install

Preparing a Debian host, the packages ProxDock needs, and the first sign-in.

  • Supported platforms
  • Install
  • Knobs
  • Or do it by hand
  • After it installs
  • +3 more

First run

The setup wizard: creating the cluster, defining network zones and mapping them to interfaces.

  • What the installer left you
  • 1. Unlock
  • 2. Create or join
  • 3. Name the cluster and give it storage
  • Creating a pool from a blank disk
  • +6 more

Compose files

What the transformation engine changes, what it refuses, and the variables a file may refer to.

  • What gets changed
  • What gets refused
  • Referring to your own identity
  • Reading the preview

Backups and restore

Send streams rather than tarballs, scheduled jobs, retention, and restoring under a new name.

  • What a backup run does
  • Restoring
  • Browsing without restoring
  • Scheduling and retention
  • Remote targets

Running a cluster

Joining nodes, what quorum protects, and migrating an application between hosts.

  • How a request is served
  • Joining a node
  • Quorum
  • Migrating an application
  • Upgrading

Reference

Architecture

How the pieces fit: the store, the orchestrator, ZFS, and where each decision is made.

  • The pieces
  • Two principles worth stating
  • Where decisions are made

Command line

The binary's small set of host-level subcommands, and why deploying is not one of them.

  • Subcommands
  • Flags
  • Getting back in
  • Scripting against ProxDock

Certificates

Per-node HTTPS over ACME DNS-01, and using certificates you manage yourself.

  • ACME over DNS-01
  • Certificates you manage yourself
  • Certificates for applications