Fifteen years of databases, end to end
Fifteen years of database platforms at AWS, Oracle, Tessell, OpenSCG, and EnterpriseDB — and how they became CortexBolt, the autonomous operations platform.
Dinesh4 min read
The CortexRidge Blog
Database operations, deeply explained.
Working notes from the operational layer. PostgreSQL, ClickHouse, MySQL, MongoDB — and what it takes to run them at production scale. No listicles. No filler.
Featured
Latest articles
PostgreSQL8 min
SQL vs PL/pgSQL in PostgreSQL: Plan Caching
SQL vs PL/pgSQL functions in PostgreSQL 16: why a SQL function is re-planned on every loop call, how to reduce plan time.
PostgreSQL5 min
ALTER FUNCTION SET work_mem: Fixing PostgreSQL Disk Spills Without a Global Change
Use PostgreSQL ALTER FUNCTION SET work_mem to stop one function's sorts spilling to disk without a global change. A production fix: 150 GB a day to 0.
PostgreSQL5 min
pgsonify: Hearing PostgreSQL Health as Elephant Sounds
pgsonify is an experimental MIT-licensed tool that sonifies PostgreSQL health metrics as real elephant recordings. How it maps stats views to sound.
Database Operations4 min
Workflows in Practice: Disk Space Monitoring
One webhook call checks a server's disk, records the reading, alerts the customer, and archives the run. A five-node CortexBolt workflow, step by step.
PostgreSQL16 min
The Cost of Storing and Indexing Embeddings in pgvector
What does it cost to store and index embeddings in pgvector? Four costs priced per vector, and why keeping the HNSW index in RAM is the one that sets the bill.
PostgreSQL8 min
PostgreSQL LISTEN/NOTIFY Is Not Durable — Here's the Fix with pg_durable
PostgreSQL LISTEN/NOTIFY loses messages when nobody is listening. See how pg_durable adds durable signals and workflows that survive restarts.