All Products
Search
Document Center

Lindorm:Manage the vehicle route data for a ride-hailing platform

Last Updated:Mar 28, 2026

A ride-hailing platform with 70,000+ vehicles—each uploading route and location data every minute—generates roughly 100 GB of new data per day, accumulating to over 100 TB across three years. At that scale, on-premises MySQL databases struggle to keep up: query response times degrade and storage costs climb. Migrating to Lindorm resolves both problems without changing application code.

Challenges

  • Slow queries from MySQL overload: With 70,000+ vehicles continuously writing data, on-premises MySQL databases become overloaded, causing online queries to slow down significantly.

  • High storage costs at 100 TB scale: Three years of retention at ~100 GB/day results in over 100 TB of total data, making on-premises storage expensive to maintain.

Solution

Architecture diagram showing data flow from vehicles through Kafka/Spark to Lindorm with OSS cold storage

The platform migrates all historical data from MySQL to Lindorm and uses Apache Kafka or Apache Spark to sync incremental data to Lindorm in real time. This reduces the write load on on-premises databases while keeping the application layer unchanged.

Tiered storage for hot and cold data

Only the most recent month of vehicle data is accessed frequently. Lindorm automatically separates hot and cold data based on age, with no code changes or manual data management required:

TierData ageStorageAccess pattern
HotCurrent monthUltra disksLow-latency online queries
ColdOlder than one monthObject Storage Service (OSS) bucketsInfrequent; still queryable

Data older than one month is automatically migrated to OSS at lower cost. The separation is transparent to the platform.

Operational queries with SQL script templates

Operations staff can run routine tasks—such as locating a vehicle—using SQL script templates provided by Lindorm.

Benefits

  • Lower storage costs: Storing Internet of Vehicles (IoV) data in Lindorm with tiered OSS cold storage reduces costs compared to keeping all data on high-performance on-premises disks.

  • Scales with fleet growth: Lindorm's distributed storage handles growing data volumes without requiring architectural changes as the number of vehicles increases.