Tag
Pydantic 的技术博客介绍 Logfire 通过 dynamic shredding（动态拆解半结构化 JSON 属性数据）技术实现部分查询提速 1000 倍，原本 30 秒超时的查询现在不到一秒即可完成，将于 2026 年 2 月向所有客户推出。
GreptimeDB 通过将 TopK 运行时边界下推到扫描层（基于 DataFusion 的动态过滤器），把一个 ~5-6B 行 trace 表上的 ORDER BY ... LIMIT 10 查询从 28.9 秒优化到 0.21 秒。
A technical deep-dive blog series exploring join ordering in query optimization, focusing on the German school of dynamic programming approaches (DPsize, DPsub, DPccp, DPhyp) and related work on adaptive optimization and join search space enumeration.
The article explains that PostgreSQL's SELECT DISTINCT clause does not scale efficiently, as it always scans all matching rows, leading to performance problems in certain workloads, and provides insights and workarounds.
A discussion about using Jev for query optimization, with one user sharing a 12% speed improvement in Postgres queries and another commenting on Jev's performance issues.
The article describes an experiment where a 4B AI model is trained via fine-tuning and reinforcement learning to optimize Postgres query plans, achieving an 81% faster performance and 44.7% latency reduction across join-heavy queries.
QueryBrew is a proposed system-agnostic approach for optimizing SQL-to-SQL queries, aiming to improve performance across different database systems.
Cloudflare improved query performance by 35x on billion-row Postgres tables using TimescaleDB (Tiger Cloud), and built a real-time earthquake dashboard with Claude Code and Tiger CLI.
This article demonstrates how to detect full table scans in SQLite using its statement statistics API, and suggests integrating this into Rails to warn or raise in test/development environments.
This article demonstrates clever techniques to achieve partition pruning in PostgreSQL when filtering by non-partition key columns, going beyond conventional wisdom.
PlanetScale introduces Database Traffic Control, a Postgres traffic management system that lets users enforce flexible budgets on database resources, preventing overload from bad queries or runaway workloads.
pg_plan_advice is a PostgreSQL module that allows users to influence and stabilize query plan choices by specifying plan advice using a mini-language. It helps in controlling join order, scan methods, and other planner decisions, but overriding planner defaults can backfire if data distribution changes.
This article explains how continuation-passing style (CPS) can be used to fuse database operators, avoiding materialization of intermediate results and improving performance, as an alternative to vectorization and compilation. The author presents a simplified example with numeric list operations and describes the inspiration from a short paper.
Postgres 19 introduces query hints via new contrib modules pg_plan_advice and pg_stash_advice, ending a long-standing community debate and providing DBAs with an escape hatch for optimizer edge cases.
PostHog used an AI agent based on Karpathy's autoresearch to find a three-year-old bug in their ClickHouse query engine that prevented proper primary key usage for timestamp filters. Fixing it improved performance by 11% and reduced scanned granules by 62%.
GreptimeDB v1.0 integrates DataFusion's dynamic filter pushdown to accelerate TopK queries by pushing runtime bounds to the scan layer, reducing query time from 29 seconds to 0.21 seconds on a 5B-row trace table.
A tutorial on optimizing MongoDB query performance using indexes, demonstrating how to identify slow queries, apply compound indexes, and manage them visually using the VisuaLeaf tool. The workflow covers query profiling, index recommendations, and common indexing mistakes.