A key challenge in a distributed SQL database is achieving high throughput and consistency for both write and read operations. This blog explores the new YugabyteDB Cost Based Optimizer (CBO), an essential piece of the puzzle to achieve high throughput for read operations. It highlights the challenges we tackled while developing the CBO, and details how some of the most impactful aspects of YugabyteDB are modeled.
Delivering query parallelism is a key step on our journey to narrow the performance gap between Postgres and YugabyteDB. This blog examines the Postgres Parallel Query feature supported and available in the latest version of YugabyteDB. Click to discover how query parallelism is implemented and functions in Postgres, and what we did to ensure the feature works in YugabyteDB.
A key part of our effort to achieve full Postgres compatibility is providing support for bitmap scans (a scan type that combines multiple indexes). This blog deep dives into the high-level strategy of bitmap scans and examines how Postgres implements bitmap scans, how we adapted bitmap scans to work for YugabyteDB, and areas where YugabyteDB bitmap scans go beyond vanilla Postgres.
We aim to prove YugabyteDB’s enhanced Postgres compatibility by migrating 100 popular open source Postgres applications to YugabyteDB! Discover how you can get involved in the Postgres App Century Challenge.
Conducting extensive and diverse performance and scalability testing ensures that YugabyteDB remains a reliable and high-performance distributed PostgreSQL database, ideal for handling mission-critical, global-scale, applications. This blog gives a deep dive into our comprehensive testing framework.
Geo-partitioning with YugabyteDB (versus standing up separate PostgreSQL instances per region) is the ideal solution for use cases that require cross-region data access, faster time to market, and easy deployment management.
Read this blog to discover the pitfalls of standing up separate PostgreSQL instances per region and the advantages of using geo-partitioning in YugabyteDB.
Developers can now store, index and query vectors in PostgreSQL using pgvector, eliminating the need for a dedicated vector database. This blog deep dives into pgvector, and explains how to store text embeddings as vectors, execute similarity search, choose between indexes, and scale with distributed SQL.
Postgres is increasingly being used for Gen AI workloads, but users have reported scalability and performance issues, data privacy challenges and high availability concerns. Before you choose a scalable, HA, vector database instead, try running distributed PostgreSQL. Read our new blog to discover when and how to use distributed PostgreSQL for Gen AI workloads.
Discover why YugabyteDB is moving beyond PostgreSQL 11 and the challenges and solutions involved in upgrading to PostgreSQL 15 (and beyond). Learn how our engineering team is bringing the most popular PostgreSQL features to YugabyteDB, to deliver a reliable, scalable, highly available, and secure distributed SQL database.
PgBench is widely used for testing database performance, but its default ‘TPC-B like’ workload may not accurately represent true performance. To achieve more realistic load testing, users can employ custom SQL scripts within PgBench. In this demonstration, we’ll show how these adjustments work using YugabyteDB.
