Grab Migrated Counter Service Backend to Aerospike
The migration has significantly boosted system performance while cutting storage and operational costs.
Updated on Oct. 7, 2026 in Data Centers
Live Poll
Do you trust that complex technical upgrades by large platforms lead to better services for you?
Grab successfully transitioned its Counter Service storage backend to Aerospike, achieving a 50 percent reduction in p99 read latency. The massive platform, which manages one billion daily requests, completed this transition without any service downtime.
Why it matters
By moving away from a wide-column database that required multiple network round trips, Grab successfully optimized its anti-fraud infrastructure for speed and efficiency. This shift demonstrates a strategic move to lower operational overhead while handling high-volume data demands.
The transition reduced on-disk data storage from 3 TB to 1 TB, while record counts dropped by more than an order of magnitude. The system now utilizes 64 bytes of primary index metadata per Aerospike record.
The players
Grab
Grab is a leading Southeast Asian superapp company that provides ride-hailing, food delivery, and digital financial services.
Aerospike
Aerospike is a high-performance, distributed NoSQL database platform designed to handle real-time data at scale.
The details
To execute the migration, Grab implemented a storage facade to decouple business logic from storage access, allowing for seamless traffic shifts. The engineering team utilized shadow reads and gradual traffic routing to ensure the transition occurred without interrupting the Counter Service.
Timeline
October 7, 2026: The migration results were publicly detailed.
The Tech Race
This migration reflects a broader industry trend of re-evaluating legacy NoSQL architectures in favor of specialized, high-performance distributed databases. By replacing designs that suffer from high network round-trip overhead, Grab positions its infrastructure to scale more efficiently against competitors.
Users of the platform will experience improved reliability and faster response times in features protected by the anti-fraud system. The increased efficiency allows the service to maintain high performance even as transaction volume continues to scale globally.
The takeaway
Optimizing backend storage can yield massive performance gains without requiring a total system overhaul. Companies dealing with massive request volumes can achieve significant cost savings and latency reductions by migrating to specialized, high-performance database architectures.
Further reading
Learn more about the latest innovations in high-scale infrastructure within our Data Centers section.
Source note: This article includes information reported by InfoQ.
Live Poll
Do you trust that complex technical upgrades by large platforms lead to better services for you?
