Case Study: Fastenal achieves sub-second product search and faster nightly index refreshes with Elastic

Fastenal - Customer Case Study

Fastenal is a global industrial distributor with thousands of branches, onsite locations and millions of SKUs supporting heavy e‑commerce and enterprise use. Their challenge was keeping product search fast and current: the original Elasticsearch product index grew to ~1.8 TB, produced unbalanced shards and required >24 hours to rebuild, causing slow search responses and making nightly refreshes impractical at scale.

Fastenal rebuilt the indexing pipeline and data model: they denormalized catalog data into product documents, eliminated expensive parent/child structures where possible, and introduced a Gearman-driven PHP index builder with worker pools plus a "helper" index to perform single-pass updates via scrolling. The result was a primary product index reduced to ~375 GB, balanced ~30 GB shards, index builds cut to under 12 hours, search response improved to <1s for hundreds of users, and reliable support for millions of daily API requests.

Fastenal

Nicole Albee

IT DevOps Engineer