redb.Tsak

runtime container

The .NET analogue of Apache Camel K

The runtime container for redb.Route integration pipelines. Drop DLLs into a folder; start, stop and hot-reload them at runtime. Scale across nodes with leader election. Watch every message, error and restart — from a REST API, a 30-command CLI, or a Blazor dashboard. Zero downtime, production-tested. The missing piece between a hand-rolled Worker Service and a heavyweight enterprise ESB.

GitHub ↗ NuGet ↗llms.txt ↗
351
tests passing
32
REST endpoints
30
CLI commands
Blazor
dashboard

What's inside

  • Module-based deployment: drop a .dll or .tpkg into Libs/ — Tsak loads it and hot-swaps updates without dropping in-flight messages; per-module AssemblyLoadContext isolation.
  • Three deployment modes: standalone (in-memory) · single-node + redb (durable RTTI-storage state) · cluster (leader election + automatic context redistribution).
  • Three API modes (full / headless / readonly) + 5-layer deep-merge configuration that hot-reloads — no code changes, no restarts.
  • Cluster: epoch-fenced leader election, node registry with heartbeats, rolling hot-reload; pluggable backends (Kubernetes / etcd / Redis / Consul).
  • Observability: per-process & per-route metrics, ring-buffer logs, OpenTelemetry, optional Prometheus; watchdog auto-restarts hung routes; Quartz scheduler in every context.
  • Operate it: 32-endpoint REST API + typed C# client, 30-command CLI, 10-page Blazor dashboard. Security: API key + HMAC-SHA256, roles, revocation, redb-backed key store.

Latest releases

  • 3.6.0
    • Capture never persisted on Postgres. DlqService.CaptureAsync bound replayable as an int
    • Retention sweep and date-filtered queries threw on Postgres. Timestamps (occurred_at, since,
  • 3.5.1
  • 3.5.0
All releases (17) →

Articles on redb.Tsak

redb 3.6.0: a bug report that turned out to live in six providers — plus AS2/EDI and a shared port A user's report exposed a cross-conversation leak in the core query provider — 6 of 6. Also in 3.6.0: AS2/EDI, a shared Kestrel, Camel parity.
What an integration layer really costs: a TMS backend counted line by line An integration layer on WSO2 MI and EF Core, counted line by line: what a real TMS backend costs to keep running.
A TMS backend, architecturally: cluster, coordinate streams and a typed object store How a TMS backend is built: fifteen modules over three nodes, coordinate streams through Kafka and Redis, and a typed object store.
We ported a module off an ESB and the code grew 1.7×. That was the right outcome One module moved off the ESB and the code grew 1.7 times. Why that was the right outcome, counted module by module.
A payment platform on .NET: what it actually costs, and how much of it you never have to write What a payment platform on .NET actually costs to build, and how much of it is already solved by the runtime it runs on.
Reference architecture for a payment platform on .NET: the ledger, transaction boundaries, modules, and the cluster Reference architecture for a payment platform: the ledger model, transaction boundaries, the module split and cluster topology.
redb 3.4.0: day-two operations for a .NET stack — replay what failed, hot-patch the framework, lock down the control plane Day-two operations for a .NET stack: replay what failed, hot-patch the framework without a rebuild, and lock down the control plane.
redb 3.3.0: an enterprise .NET stack you actually own — typed store, a homegrown Apache Camel, and a runtime with a dashboard (all free) An enterprise .NET stack you actually own: a typed store, a homegrown Apache Camel and a runtime with a dashboard.
All articles on redb.Tsak (9) →

Docs and examples for this product are coming soon. For now, see GitHub and the README.

← Back to all products