Location Services cost optimization

Context

Location Services was developing a major system that’s hosted in Azure, utilizing various tool and services.

Challenge

The ongoing monthly hosting costs were increasing significantly, causing concerns.

Solution

We’ve analyzed Azure spending, and identified CosmosDB as a service that contributed to the increased costs.

We proceeded to investigate CosmosDB usage and query patterns. We noticed a few heavey queries that were consuming a varge large number of RU/s.

We advised development team on how to modify the code and optimize the queries, which resulted in a dramatic decrease of RU/s used. This allowed the team to reduce the provisioned throughput for CosmosDB on multiple environments and the final cost savings were over $100,000 per year.