As ServiceNow continues to scale for enterprise-grade workloads, the traditional database approach (MariaDB) started showing limits – especially in handling massive amounts of transactional and reporting queries simultaneously. With organizations generating billions of records across ITSM, IRM, HRSD, and other modules, performance bottlenecks became a serious challenge. Queries that needed instant results were slowed down by sequential scans and limited indexing capabilities.
This is where RaptorDB, ServiceNow’s new database engine, comes into play. RaptorDB focuses on speed, scalability, and intelligence in how data is indexed, stored, and retrieved. By combining smarter indexing, parallel processing, and modern storage techniques, it makes queries up to 10x faster and sets the stage for AI-driven operations.
What Is RaptorDB?
RaptorDB is ServiceNow’s high-performance distributed document database designed for large-scale analytics, telemetry, and AI-driven workloads. Unlike the relational MySQL database that powers transactional operations on the Now Platform, RaptorDB is optimized for speed, scale, and flexibility when dealing with billions of records and real-time data processing.
Why Was RaptorDB Needed?
ServiceNow’s traditional relational database (MySQL) is best suited for transactional workloads (e.g., incident, change, request records). However, enterprises increasingly require:
- Real-time insights from large datasets.
- High-speed search and analytics over billions of records.
- Scalable infrastructure to support AI-driven decision-making.
How Does RaptorDB Work?
Traditional databases like MariaDB were built mainly for storing and retrieving records, but they slow down when handling heavy analytics or large queries. RaptorDB is designed to do both at once – transactions and analytics – on the same platform.
Think of a library analogy:
- With MariaDB, you have one librarian searching shelf by shelf until the right book is found.
- With RaptorDB, you have multiple librarians using a catalog/index that instantly tells them where each book is stored. They fetch results in parallel.
This means queries are resolved faster, even when millions of “books” (records) exist.
Indexing in RaptorDB
Indexing is at the heart of RaptorDB’s speed advantage. Instead of scanning rows one by one, RaptorDB builds smarter indexes that act like a map to your data. This allows the database to jump directly to the right information, even across massive datasets, and process queries in parallel.
Indexing is the secret sauce:
- MariaDB indexing: Like a single index card, good for small searches but slower for massive, complex queries.
- RaptorDB indexing: Uses multi-dimensional indexes that can search across multiple fields simultaneously.
- Dynamic updates: Indexes adjust automatically as data changes, ensuring freshness.
- Parallel lookups: Instead of one thread, multiple processors work together, cutting down query time.
Why it’s faster: You’re not scanning data row by row. You’re jumping directly to the results using a “map” of where they live.
Architecture Behind the Scenes
Traditional databases like MariaDB follow a sequential process to handle queries, while RaptorDB uses smart indexing and parallel processing to break tasks into smaller chunks and run them simultaneously. This architectural shift is what makes complex queries and analytics run dramatically faster.
Here’s how the architecture shift looks:
Old Way (MariaDB)
New Way (RaptorDB)
Are You on RaptorDB?
Follow the steps below to check if your instance is on RaptorDB:
- Go to https://YOURINSTANCE.service-now.com/replication.do
- Check the JDBC URL field.
- If
jdbc:mysql://… , then you are still on MariaDB. - If
jdbc:postgresql://… , then you are are on RaptorDB.
RaptorDB Standard vs. RaptorDB Professional
- Standard: RaptorDB Standard is the target state default primary database, built to provide improved scale and performance capabilities for ServiceNow AI Platform workloads and queries.
- Professional: If your organization is looking to drive ultra-scale and high performance to support ServiceNow AI Platform generative AI and machine-scale data use cases, user growth, applications, workloads, and analytics, you may wish to leverage RaptorDB Professional. RaptorDB Professional is a licensed database; for more information, contact your account team.
| Feature | RaptorDB Standard | RaptorDB Professional |
| Optimized for ServiceNow workloads | Yes | Yes |
| Improved response time | Yes | Yes |
| Ultra-scale performance and scalability | No | Yes |
| Enhanced analytics with built-in column store | No | Yes |
| Advanced instance topology support | No | Yes |
| Support for self-hosted environments | No | Yes (starting with the Zurich release) |
Summary
RaptorDB is more than just a database replacement – it’s a strategic shift in how ServiceNow handles data. By combining smart indexing, parallel processing, and hybrid transactional-analytical capabilities, it delivers faster data retrieval, scalable growth, and stronger support for AI-driven workflows.
Think of it as moving from a small library with one librarian (MariaDB) to a modern library with a digital catalog and multiple librarians working in parallel (RaptorDB). Instead of waiting while one person searches shelf by shelf, your requests are routed instantly, and results arrive in a fraction of the time.
For enterprises, this means quicker decision-making, smoother user experiences, and future-ready workflows – ensuring ServiceNow remains a powerful platform as data volumes and business demands continue to grow.