Skip to main content

Roadmap

This page explains what the project is trying to become next. It is intentionally separate from the usage docs so readers can distinguish current behavior from future design.

How To Read This Roadmap

The project is being built in layers:

  1. harden the current broker behavior
  2. strengthen storage and durability
  3. add real distributed-system behavior
  4. add production operations and safety

Major Phases

PhaseGoal
Prototype hardeningstronger parsing, configuration, logging, and baseline tests
Storage engine v1segmented logs, indexes, retention, recovery, durability knobs
Replication and leadershipfollower replication, ISR, high watermark, ack modes
Consumer coordination v2heartbeats, rebalances, generation-aware commits
Metadata quorumcontroller service, leader election, broker registration, fencing
Security and multi-tenancyTLS, auth, ACLs, quotas
Observability and operationsmetrics, traces, admin tooling, runbooks
Performance validationbenchmarks, soak tests, chaos tests, optimization

Near-Term Work

The immediate engineering focus is on the pieces that make the current product safer to evolve:

  • explicit protocol envelopes and error codes
  • configuration validation
  • segmented log storage
  • broader unit-test coverage
  • CI automation
  • initial metrics support

What “Production-Ready” Would Mean

The project should only claim production readiness after it demonstrates:

  • durable storage with recovery guarantees
  • proven replication behavior under failure
  • operational visibility and runbooks
  • secure defaults
  • repeatable performance targets

Until then, the product should be described as an early-stage streaming broker rather than a production-grade distributed platform.

For the full execution plan, see the production plan.