Digital commerce platform migrates terabytes of sales analytics data from self-hosted Graylog to Elastic Cloud 8.3 in 8 weeks
Client
The client is an international digital commerce agency that powers over 45 global fashion brands, enabling scalable e-commerce operations with real-time analytics and storefront reliability.
Project Context
The client’s Graylog-based system (on Elasticsearch 2.x) had become outdated and unstable. Migration to Elastic Cloud 8.3 was required to ensure high availability during peak holiday seasons,without impacting storefront operations.
Project Objectives
Migrate 2.4 TB of analytics data across 9000 indices and 2.2B documents.
Ensure zero downtime and no impact on live storefronts.
Train the LiveOps team post-migration.
Challenges
Legacy stack with limited scalability.
No existing in-house Elastic expertise.
High-pressure timeline aligned with upcoming sales events.
Solution
Designed a multi-step upgrade path (2.x → 5.x → 7.x → 8.3).
Used AWS Load Balancer and AWS SQS to buffer and ingest GELF logs as JSON.
Enabled live data capture during migration and ensured a smooth switchover.
Solution Delivery
SquareShift architected and executed a hybrid migration strategy. Historical and live data were separated, staged, and streamed using AWS SQS and custom ingestion scripts. Post-deployment, the client's teams were trained for Elastic Cloud ops.
Testimonial
SquareShift’s migration gave us a cloud-native analytics foundation - fast, reliable, and seamless.
