Migrating Graph from a (MariaDB & MongoDB) solution to Azure Cosmos DB's Gremlin API
summary
Migration moves a fraud-analytics graph from MariaDB (link tables) + MongoDB into Azure Cosmos DB Gremlin API and reports concrete results: a dataset of 2 billion vertices and 4 billion edges, write load peaking 5,000/sec (14M/day), and point reads at ~300 TPS. Key choices: synthetic partition key to balance writes and reads, limit traversals to 1–3 hops, measure RUs with the execution profile, test the top 25 queries and tune to reduce request charge, and plan for the 20 GB logical partition limit and occasional Spark offload for large batch workloads.
AI-written summary · about
chapters (5)
- 0:00Introduction
- 0:53Use Case
- 6:10Need for Modernization
- 9:22Solution Recommendation
- 26:54Tips and Guidance
related talks
57mGraphing Your Way Through the Cosmos: Common Data Problems Solved with Graphs
01h 01mFrom SQL to Azure Cosmos DB
13mConnecting the Dots: The Path from Document Database to Graph Database
45mGetting the Most from NoSQL in the Cloud with Azure Cosmos DB
01h 02mFrom SQL to Azure Cosmos DB
49mModern Graph Databases
01h 00mWho’s Afraid of Graphs?
54mSay Yes to NoSQL for the .NET SQL Developer