SNCF: GOC 2.0 architecture audit

CASE STUDY · TRANSPORT & MOBILITY

Audit, Dynatrace instrumentation and Root Cause Analysis for national rail traffic.

SNCF called on Adservio to audit the GOC 2.0 architecture, instrument its 82 services with Dynatrace + OpenTelemetry and deploy a continuous RCA approach.

SNCF : Adservio case study
Client
SNCF (national rail traffic)
Expertise
Architecture Audit · Dynatrace Observability · Root Cause Analysis
Tech Stack
Dynatrace · OpenTelemetry · TOGAF · ArchiMate
Engagement
8 experts · 18 months
CONTEXT

Project Context

SNCF's GOC 2.0 programme sits at the heart of the operational management of national rail traffic. Any failure has a direct impact on passengers, freight and the safety of the network.

The architecture, built up over years, had operational and technical blind spots. Incidents took several days to diagnose, due to a lack of end-to-end observability across the 82 services.

The stake: audit the architecture according to TOGAF, instrument via Dynatrace + OpenTelemetry, and put a continuous RCA in place, without ever disrupting rail traffic or degrading operational SLAs.

Strategic Objectives

(01)

Critical architecture audit

Audit the architecture of the GOC 2.0 programme (operational traffic management) according to TOGAF, identify critical risks and propose a prioritised action plan.

(02)

End-to-end observability

Instrument the 82 application services with Dynatrace + OpenTelemetry to obtain complete visibility over data flows, traces, logs and production metrics.

(03)

Root Cause Analysis

Set up a continuous RCA capability combining Dynatrace traces and architecture expertise, to resolve critical rail traffic incidents in a lasting way.

Solutions Delivered by Adservio

Adservio deployed an Architecture + Observability team (TOGAF Architect, SRE Lead, Dynatrace experts, RCA specialist) over 18 months.

(01)

Complete TOGAF audit

A methodical audit of the GOC architecture according to TOGAF: data, applications and infrastructure layers. ArchiMate mapping, identification of major risks and a remediation roadmap.

(02)

Dynatrace instrumentation

Deployment of Dynatrace OneAgent across the 82 services, configuration of dashboards by business journey (planning, signalling, traffic), and setup of AIOps Davis for automatic anomaly detection.

(03)

OpenTelemetry traces

Standardisation of traces on OpenTelemetry for interoperability, end-to-end context propagation and intelligent sampling on high-throughput services.

(04)

Continuous RCA

A continuous RCA capability: systematic post-incident analyses, correlation of traces, logs and metrics, blameless postmortems, and the creation of a library of incident patterns to speed up diagnosis.

(05)

Prioritised action plan

An action plan validated by the SNCF technical management, structured by layers (data 12 TB/day, applications 82 services, infrastructure on-prem and cloud), prioritised by business criticality.

82 services, 12 TB of traces per day,
a continuous RCA.

Complete TOGAF audit, Dynatrace + OpenTelemetry instrumentation across the 82 services, and a continuous RCA approach: observability as a lever of resilience for national rail traffic.

Results

82
Services instrumented

Application services instrumented with Dynatrace + OpenTelemetry.

12 TB/day
Of traces processed

Volume of time-series traces processed each day on the GOC 2.0 programme.

Days → hours
RCA time

Average Root Cause Analysis time on critical incidents.

0
SLA degradation

National rail traffic maintained with no degradation of SLAs.

3
Layers audited

Data, applications and infrastructure covered by the TOGAF audit.

18 months
Of programme

Duration of the audit, instrumentation and RCA roll-out.

Impact

Validated action plan

A complete action plan validated by the SNCF technical management, covering the 3 layers (data, applications, infrastructure) and prioritised by impact on national rail traffic.

End-to-end observability

82 services instrumented with Dynatrace + OpenTelemetry, complete visibility over data flows and traces: no more blind spots in diagnosis.

Continuous RCA

An operational RCA capability with automatic correlation of traces, logs and metrics. The average RCA time drops from several days to a few hours on critical incidents.

Anticipated anomalies

AIOps Davis identifies subtle anomalies before they become incidents: proactive detection of degradations that would previously have been invisible.

Durable architecture

TOGAF recommendations implemented progressively, transforming the GOC architecture to make it more resilient and easier to evolve over time.

Traffic maintained

Throughout the programme, national rail traffic is maintained without any degradation of SLAs: the audit and RCA do not affect operations.

TALK TO AN EXPERT

Strengthen the resilience of your critical systems

Let's discuss your architecture audit, observability and RCA challenges. An Adservio expert gets back to you within 24h.

By submitting this form, you agree to our privacy policy.