Architecture · 1970–now · Cloud Scale
Modular monolith
One deployable with clear boundaries. The unfashionable default that keeps shipping while microservices argue about ownership.
Why Modular monolith stuck
Monoliths stuck because most products die from coordination overhead, not single-process limits. Clear modules and good tests beat forty repos with stale contracts. Fashion hated it; revenue kept choosing it.
Case studies
- Big-bang microservices cutover weekend — Weekend cutover rollback; months of dual-run and incident debt
- Twitter outgrows pure Rails — without declaring Rails dead — Years of JVM/Scala investment alongside Ruby; Fail Whale era ops pain, not a total rewrite bill
- Digg v4 rewrites the product and the audience — Traffic collapse and sale; rewrite consumed a year of product oxygen
- Mid-market CRUD shop splits into forty repos — 3× infra spend; velocity flat for two years
- Service mesh before the traffic arrived — Platform team of six; product incidents from mesh misconfig outpaced actual load issues
Patterns
Compare with
Related
Architecture · 2012–2023
Microservices Everywhere
Independent deployability as religion. Correct for some; ruinous cargo-cult for most. Consolidation and monolith return are the 2020s epilogue.
$ Teams split 4-person products into 40 services, then hired SREs to babysit the mesh. Many later consolidated — after years of distributed tracing invoices and on-call trauma. See case studies on premature microservices and big-bang cutovers for named failure modes.
Architecture · 2004–now
MVC web frameworks
Model-view-controller as the default web shape. Still how most server-rendered apps think.