Rapid Recovery of a Critical IT Service Platform

THE CHALLENGE

A critical platform in need of stabilisation

After consolidating multiple legacy systems into a single unified platform, the client experienced severe instability. The new stack was oversized, under-tested and poorly integrated, leading to repeated failures, 243 hours of lost trading time per week, and escalating customer impact.

Key issues included:

– System instability and performance bottlenecks

– Excessive processing overheads and untested requirements

– Poorly defined error handling and fragile component interaction

THE SOLUTION

Fast response. Sustained performance. Long-term stability

Catapult deployed its Resolve team, leading two parallel workstreams to stabilise, simplify and recover platform performance at speed.

Workstream 1. Stabilisation & Performance Optimisation

· Migrated Oracle databases to faster SAN storage

· Accelerated system startup and shutdown, reducing release cycle times

· Implemented platform enhancements to boost responsiveness

Workstream 2. Simplification & Right-Sizing

· Reorganised server architecture to improve uptime and efficiency

· Reduced unnecessary communication “hops” between stack components

· Introduced defined error scenarios and robust testing to improve resilience

THE RESULTS

Major gains in stability, speed and efficiency

Catapult’s engagement delivered dramatic improvements in system stability and performance restoring lost trading hours, accelerating delivery cycles and significantly reducing operational overhead.

– 243 trading hours recovered per week

– Message failures dropped from 12% to just 0.3%

– Billing system performance increased by over 1000x

– CRM and stack performance improved by 30%

– Useful database processing jumped from <50% to 90%

– Release outage window reduced from 21 days to just 30 minutes

– 42 physical machines freed up, boosting availability and reducing cost