Skip to content

BLOG-BACKLOG.md

Latest commit

 

History

History
74 lines (59 loc) · 5.25 KB

BLOG-BACKLOG.md

File metadata and controls

74 lines (59 loc) · 5.25 KB
Raw

Blog backlog — problem/solution posts (proposed)

Top-of-funnel, problem/solution angle (not comparisons). Each targets a search a developer runs when they have the problem but haven't heard of @imqueue yet — the post leads with the problem, then introduces the framework as one answer. Kept on the local blog-draft branch. All 10 below are now written into src/org/blog/posts/ (drafted 2026-07; review before publishing). The "also-ran" ideas at the bottom remain unwritten suggestions.

Guidance applied to all: keep the transport framing generic ("message queue"), don't position @imqueue as tied to a single broker, no product-vs-product comparisons.

  1. Stop hand-writing and hand-maintaining microservice clients

    • Target keywords: "generate typed api client typescript", "stop writing api clients", "sdk maintenance microservices"
    • Pain: every service needs a client/SDK, and they rot out of sync with the service.
    • Angle: the cost of hand-maintained clients (drift, duplicated types, release coupling) → self-describing services + generated typed clients regenerated on change.
    • Hook: imq client generate, compile-time breakage when a signature changes.

  2. Load balancing internal microservices without running a load balancer

    • Target keywords: "microservice load balancing nodejs", "load balance internal services", "load balancing without load balancer"
    • Pain: teams stand up and operate LBs/ingress just so internal services can share traffic.
    • Angle: competing-consumers pattern — N instances read the same queue, work self-balances, nothing to operate in the request path.

  3. Do Node.js back-ends really need service discovery?

    • Target keywords: "nodejs service discovery alternative", "do i need consul", "service discovery overkill"
    • Pain: Consul/etcd/DNS-SD setup and operation for modest fleets.
    • Angle: when discovery is genuinely needed vs when "the queue name is the address" removes the problem entirely.

  4. Cutting boilerplate out of Node.js microservices

    • Target keywords: "reduce microservice boilerplate nodejs", "microservice scaffolding typescript"
    • Pain: repetitive scaffolding, wiring, and client code per service.
    • Angle: what boilerplate actually accretes in a service repo → scaffolding + generated clients + conventions that remove it. Show a before/after file count.

  5. Back-pressure for Node.js services (and why HTTP makes it hard)

    • Target keywords: "nodejs backpressure microservices", "handle traffic spikes microservices", "overload cascading failure"
    • Pain: a slow/downstream service takes callers down with it under load.
    • Angle: how a queue absorbs spikes instead of dropping connections; where guaranteed delivery fits; the trade-offs to watch (unbounded growth, TTLs).

  6. Guaranteed message delivery: what it costs and when you need it

    • Target keywords: "guaranteed message delivery nodejs", "at-least-once delivery", "reliable messaging"
    • Pain: "will I lose messages if a worker dies?" and fear of the performance cost.
    • Angle: unreliable vs safe delivery, the roughly-halved-throughput trade, and a decision guide per workload. (Grounded in the real delivery-mode behavior.)

  7. Testing services that call each other — without booting the whole stack

    • Target keywords: "test microservices nodejs", "mock service calls typescript", "integration test microservices"
    • Pain: end-to-end tests need every service + infra running; slow and flaky.
    • Angle: strategies for testing across service boundaries — typed clients as seams, testing a service in isolation, when to use a real queue vs a fake.

  8. Versioning microservices without breaking every caller

    • Target keywords: "microservice versioning nodejs", "breaking change microservices", "coordinate service releases"
    • Pain: a signature change silently breaks callers; coordinating releases is manual and scary.
    • Angle: regenerate-client workflow surfaces breakage at compile time; coordinated version bumps across a fleet; additive-change discipline.

  9. From monolith to services: a low-risk first extraction for Node.js teams

    • Target keywords: "monolith to microservices nodejs", "extract first microservice", "strangler fig nodejs"
    • Pain: big-bang rewrites are risky; teams don't know where to start.
    • Angle: extract one bounded capability behind a typed service call, keep the rest in place, iterate. A concrete step-by-step.

  10. Internal APIs don't need to be REST

    • Target keywords: "internal api rpc", "rest for internal services", "service to service calls nodejs"
    • Pain: REST ceremony (routes, verbs, status codes, serializers) for calls that are really just "run this function over there."
    • Angle: for internal traffic, function-call semantics beat resource semantics; what you gain (types, less mapping) and what REST is still better at (public/edge APIs). (Concept post, not a product comparison.)

Also-ran ideas (if you want more later)

  • Delayed & scheduled work without adding a separate job system
  • Graceful shutdown and zero-drop deploys for queue-based services
  • Structured logging and request tracing across service calls
  • Typed events (not just request/response) between services
  • Running a local multi-service fleet on one laptop for development