Four Critical Questions to Ask Before Trusting a 4x Proxy Benchmark
A ClickHouse engineering post claiming 4x throughput gains for PgBouncer attracted significant attention on Hacker News in July 2026, prompting a closer look at how such benchmarks should be…
Read the full articleYou might also wanna read
Scaling PgBouncer across multiple cores with so_reuseport in ClickHouse Managed Postgres
PgBouncer is single-threaded, so a single process caps out at one CPU core no matter the box size. See how ClickHouse Managed Postgres runs
ClickHouse·5d agoScaling PgBouncer across multiple cores with so_reuseport in ClickHouse Managed Postgres
PgBouncer is single-threaded, so a single process caps out at one CPU core no matter the box size. See how ClickHouse Managed Postgres runs
clickhouse.com·5d ago
PostgreSQL on AWS EC2: Benchmarking 26 Instance Types for Cost-Effective Performance
Size a PostgreSQL server on AWS: choose your dataset size and target throughput to find the cheapest EC2 instance that clears it — backed by
postgres.saneengineer.com·9d ago
Benchmarking Postgres Write Scalability for Durable Workflow Execution
Benchmarking the workflow execution and workflow queueing scalability of a single Postgres server.
dbos.dev·2mo ago
Performance Benchmark: Testing HTTP Request Capacity of a Single Machine Architecture
When designing systems and deciding on the architecture, I often hear justifying the use of microservices and other complex solutions becaus
binaryigor.com·10mo ago
ClickHouse Introduces PostgresBench: An Open, Reproducible Benchmark for Managed Postgres Services
Discover how Postgres managed by ClickHouse compares to its peers in a reproducible, open benchmark.
ClickHouse·26d agoClickHouse Introduces PostgresBench: An Open, Reproducible Benchmark for Managed Postgres Services
Discover how Postgres managed by ClickHouse compares to its peers in a reproducible, open benchmark.
clickhouse.com·26d ago
AI Benchmark Said It Was Better. Your Production Database Disagreed
Generic LLM benchmarks measure language model capability. They don’t measure whether your AI SRE tool can still diagnose connection pool sat
Comments
Sign in to join the conversation.
No comments yet. Be the first.