Executive Summary
A UK travel company’s high-traffic WordPress platform suffered a P1 incident caused by sustained high CPU on its RDS MySQL database. The emergency response (scaling up RDS) stabilised the platform but nearly tripled the monthly database spend and left the underlying read-amplification problem untouched. Logicata introduced an Amazon ElastiCache (Valkey) caching layer with Object Cache Pro for WordPress, absorbed the repetitive read traffic, scaled RDS back down, and delivered around $1,632 per year in net savings, with the P1 risk eliminated.
Customer Overview
Sector: Travel & Hospitality
Location: United Kingdom
The customer is a long-established UK travel company operating a high-traffic WordPress platform that serves thousands of visitors daily. Visitors browse detailed travel itineraries, destination content, and booking information. The platform runs on AWS infrastructure managed by Logicata under an ongoing managed services agreement.
The Challenge
A P1 incident exposed an architectural gap that scaling alone could not close:
- Sustained high CPU on RDS MySQL. Every page load was hitting the database directly for content that changed infrequently.
- Emergency scale-up. RDS was scaled to a larger instance class to stabilise the platform. The site recovered but the bill nearly tripled.
- Symptoms, not causes. The underlying read pressure remained. Another traffic spike could trigger the same P1 on the larger instance.
- No caching architecture. Repetitive database queries for travel descriptions, itineraries, and destination pages were hitting RDS on every single page view.
- Risk of recurrence. Without addressing the read amplification, future P1 incidents were likely as traffic grew.
Scaling the database had bought time, not a solution.
The Solution
Logicata designed and implemented an Amazon ElastiCache (Valkey) caching layer to absorb repetitive read traffic.
Technology selection
- Valkey chosen over Redis OSS as a cost-effective, fully compatible alternative: same protocol, same performance, lower price point.
- Object Cache Pro selected as the WordPress plugin to provide application-level cache control, configurable TTLs, and intelligent cache invalidation.
Staged implementation
- Staging first. Built ElastiCache Valkey in the staging environment, integrated Object Cache Pro, and let the customer’s developer validate application behaviour and cache hit rates.
- Production deployment. Built ElastiCache Valkey in production following the validated staging configuration.
- RDS right-sizing. Once the cache was confirmed to be absorbing read traffic, the RDS instance was safely scaled back down.
Network integration
- ElastiCache Valkey deployed within the existing Amazon VPC with appropriate security group rules.
- Application servers granted access to the cache endpoint with no public exposure.
Why Valkey rather than Redis
Valkey is a fork of Redis OSS that maintains protocol and performance parity, with AWS managing it as a first-class option in ElastiCache. The relevant difference here is pricing. For a workload that needs Redis semantics but not Redis Enterprise features, the cost difference is significant and there is no functional gap.
Results
- RDS CPU significantly reduced as the cache absorbed the majority of read queries
- Around $136 per month in net savings (about $1,632 per year)
- RDS right-sized back to the original instance type
- P1 incident risk eliminated. Cache prevents read pressure spikes from reaching RDS.
- Page performance improved. Cached content served without a database round-trip.
- Staging parity maintained. Both environments now have cache layers for accurate pre-production testing.
AWS Services Used
- Amazon ElastiCache (Valkey)
- Amazon RDS (MySQL)
- Amazon VPC
- Amazon EC2
- Amazon CloudWatch
About Logicata
Logicata is an AWS Advanced Partner holding the AWS Cloud Operations Management Competency, validated through an independent third-party audit. Logicata helps organisations build and operate secure, well-governed cloud platforms on AWS, enabling customers to reduce operational risk, meet assurance expectations, and scale with confidence.
