Edenic Case Study
18-Service Microservices Overhaul
4-week breakup of a monolith into 18 services delivering 12 daily releases, zero-downtime deploys, and 99.985% uptime.
Client
Mobility marketplace (white-label)
Project
Express monolith → 18 services
Timeline
4 weeks from audit to cutover
Scale
2.4M annual trips, 48K peak drivers
Deploys
12 prod releases/day, <10 min rollback
Reliability
99.985% uptime post-cutover
Context
Executive summary
A high-signal overview of the business and platform shift.
-Monolith handled every API, WebSocket, and cron; deploys risked outages and forced windows.
-Edenic decomposed into 18 services with domain-aligned ownership and templated CI/CD.
-Releases are now 12/day, rollbacks take under 10 minutes, and uptime holds at 99.985%.
What changed
Target architecture snapshot
Core platform decisions that unlocked scale and control.
Component → Target solution
Services18 Dockerized services with domain ownership
MessagingKafka for events and async workflows
DataPostgreSQL + PostGIS for OLTP; ClickHouse for analytics
CacheRedis for hot paths and session handling
StorageAmazon S3 for artifacts and assets
CI/CDTemplate pipelines for build, test, deploy per service
ObservabilityUnified metrics, logs, and SLO dashboards
Before
Core challenges
Why the platform had to change.
-One Express monolith carried every API route, WebSocket, and cron job
-Deploys took 30–40 minutes, forced maintenance windows, and risked full-site outages
-Scaling meant duplicating the whole container—costly and coarse-grained
-Sparse metrics; payment failures hid behind other traffic spikes
Intent
Migration goals
Outcomes that guided every decision.
-Break core domains (Trips, Location, Payment) into independent, horizontally scalable services
-Remove maintenance windows and enable sub-10-minute, zero-downtime deploys
-Stand up a reliable rollback path with versioned artifacts and health gates
-Add real observability across services, data, and background jobs
Execution
Phased delivery
A structured migration plan to de-risk cutover and enforce standards.
Phase 1 - Audit & Target Design
-Monolith dependency mapping across APIs, jobs, and data
-Cutover and rollback strategy with health gates
-Domain boundaries and service contracts defined
Phase 2 - Platform & CI/CD
-Templates for build/test/release with versioned artifacts
-Automated deploys with health checks and staged rollout
-Shared configs and secrets patterns standardized
Phase 3 - Service Breakup
-Trips, Location, Payment, Identity, Notifications carved out first
-Read/write segregation to reduce blast radius
-Backfills and data alignment to keep parity during migration
Phase 4 - Observability & SRE
-Centralized logging and tracing for all services
-SLOs for latency, error rate, and availability with alerts
-Runbooks for on-call and rollback drills
Phase 5 - Cutover & Hardening
-Staged traffic shifts with progressive validation
-12 daily releases with <10-minute rollback window
-Post-cutover reviews and backlog of incremental hardening
Impact
Operational metrics
The shifts that matter for release safety and reliability.
Deploy time
Before: 30–40 min + downtime
After: <10 min, no downtime
~4× faster
Release cadence
Before: Ad-hoc, gated by risk
After: 12/day with health gates
Continuous
Rollback
Before: Manual, slow, risky
After: <10 min, templated
Standardized
Reliability
Before: Frequent incident risk
After: 99.985% uptime
Stability lift
Metrics reflect public-safe highlights after stabilization.
Outcome
Value delivered
What the business now experiences.
Velocity
-12 production releases per day with predictable rollbacks
-Feature delivery decoupled by domain; teams ship independently
-Template pipelines keep new services consistent from day one
Reliability & Scale
-99.985% uptime since cutover, zero critical incidents
-Resource scaling per service instead of whole-app duplication
-Health gates prevent bad releases from reaching users
Observability & Data
-Unified metrics, logs, and traces across services and jobs
-ClickHouse analytics and Kafka event backbone for product insights
-Runbooks and SLOs guide on-call and post-incident review
Tooling
Technology stack
Core services behind the platform and data pipeline.
Node.jsGoRustPythonPostgreSQLPostGISRedisClickHouseKafkaAmazon S3
Snapshot
Engagement facts
Quick reference details.
Client
Mobility marketplace (white-label)
Project
Express monolith → 18 services
Timeline
4 weeks from audit to cutover
Scale
2.4M annual trips, 48K peak drivers
Deploys
12 prod releases/day, <10 min rollback
Reliability
99.985% uptime post-cutover
Want this level of modernization for your platform?
Edenic helps teams break up monoliths, standardize CI/CD, and launch resilient microservices with the observability and rollback paths needed for safe velocity.
Typical engagement includes audit, target architecture, phased migration, CI/CD standardization, and cutover/rollback planning.
Ask for an Edenic blueprint tailored to your current stack and release cadence.