At PGConf.EU, GitLab faced the challenge of deleting a 4 TB table in PostgreSQL with minimal impact. The table was archival, and the process involved considerations around WAL, caching, and ensuring no active reads were occurring.
Related articles
See all- PostgreSQL
Igor Suhorukov: pg_vexec one month on: ahead of ClickHouse on Parquet, close behind on MergeTree
Igor Suhorukov's pg_vexec, a vectorized database, outperforms ClickHouse on Parquet queries and nearly matches it on MergeTree. After a month of development, it shows significant speed improvements.
- Cloud
Percona Operator for MySQL 1.3.0: autopilot tuning, zero-data-loss async failover, encrypted PITR
Percona Operator for MySQL 1.3.0 introduces autopilot tuning, zero-data-loss async failover, and encrypted point-in-time recovery, enhancing MySQL management on Kubernetes.
- AI & Vector
Benchmarking Vector search in six database engines
The study evaluated vector search performance by loading 990,000 OpenAI embeddings into six databases, building an HNSW index in each, and measuring various performance metrics. This helps developers understand vector search capabilities across different databases.
- Insight for DBAs
How to Extract Data from InnoDB, MyISAM, and MyRocks Files Without a Running Server
A tool that can extract data from raw database files without a running server is highly beneficial for database administrators. It allows offline data extraction from InnoDB, MyISAM, and MyRocks files, useful when the server is unavailable or for quick data recovery.
- SQL
Igor Suhorukov: Which language should an AI agent use to search its memory: SQL or SPARQL?
A study found that using a knowledge graph with SPARQL for AI agent memory queries resulted in higher accuracy compared to SQL. The choice between SQL and SPARQL should be based on personal experiments and specific needs.
- ChromaDB
ChromaDB: Embeddings and Vector Databases in Python
Real Python introduces embeddings and vector databases with ChromaDB, explaining how relevant documents can provide context for LLMs. The supplied excerpt covers vector fundamentals and NumPy operations; the tutorial targets ChromaDB 1.5 and Python 3.12+.