Case Study: Alkemics achieves high database availability and a 45% reduction in memory consumption with Percona

Alkemics and Percona Guarantee Database High Availability and Uptime to Match Company Growth

Alkemics is a SaaS platform that connects suppliers and retailers to keep product data accurate, organized, and synchronized for better omni‑channel shopping experiences. As the company expanded into new markets, it needed to guarantee database responsiveness and high availability while addressing high memory use, excessive disk reads, duplicate indexes, and fragmented tables that threatened performance and uptime.

Percona conducted a Database Performance Audit, identified causes (Transparent Huge Pages, insufficient InnoDB buffer pool, unused indexes, fragmentation), and implemented fixes including disabling THP, increasing the buffer pool, tuning variables, removing duplicate indexes, and defragmenting tables with ongoing query review. The changes cut memory consumption by 45%, improved responsiveness and availability, and ensured the database could scale to meet Alkemics’ growth.

Alkemics

Guillaume Morin

Head of Production and Infrastructure