Newsdesk
Engineering
Data & AI
Industries
Enterprise Systems
Go-to-Market
Longform
DesignIndiaAll stories

Databases & Data

29 September 2026

63 new posts across 8 database and data engineering sources — 7 worth your time.

One week of Databases & Data, 7 stories, as published.

Postgres

Dinesh Kumar: ALTER FUNCTION SET work_mem: Fixing PostgreSQL Disk Spills Without a Global Change

This article demonstrates how to use `ALTER FUNCTION SET work_mem` in PostgreSQL to prevent specific functions from spilling to disk during sorts, avoiding a global `work_mem` change. A production fix reduced disk spills from 150 GB a day to 0.

Why it matters — Targeted `work_mem` adjustments at the function level can resolve performance issues like disk spills without impacting other queries, offering a granular optimization strategy.

Planet PostgreSQL · 29 September 2026 · Read the original →

Postgres

Chris van Eijk: PostgreSQL indexes for multi-tenant SaaS: tenant_id first

This post measures the impact of five different index strategies for multi-tenant SaaS applications in PostgreSQL 17, focusing on why `tenant_id` should come first in indexes. It also considers the effect of RLS and partitioning.

Why it matters — For multi-tenant SaaS, placing `tenant_id` first in PostgreSQL indexes is a critical optimization, significantly improving query performance and impacting the effectiveness of Row-Level Security (RLS) and partitioning.

Planet PostgreSQL · 29 September 2026 · Read the original →

Postgres

Chris van Eijk: Invoice numbers per tenant in PostgreSQL, without gaps

This article evaluates five methods for generating gap-free invoice numbers per tenant in PostgreSQL 17, including sequences, `max()+1`, `SERIALIZABLE` transactions, and counter rows. It measures their performance under load for duplicates, gaps and speed.

Why it matters — Generating gap-free sequential numbers per tenant in a concurrent environment is a common data modeling challenge, and this analysis provides measured trade-offs between different PostgreSQL approaches for reliability and performance.

Planet PostgreSQL · 29 September 2026 · Read the original →

Performance

How moving from Azure Cache for Redis to Azure Managed Redis can cut costs by 40%

This article demonstrates how migrating from Azure Cache for Redis to Azure Managed Redis can reduce monthly service costs by approximately 40%. It cites specific comparisons showing 43–44% reductions for equivalent high-availability setups in East US 2 Premium-to-Balanced scenarios.

Redis · 29 September 2026 · Read the original →

Data Engineering

How Fountain rebuilt its data plane on ClickHouse Cloud to power Cue, the Frontline Superintelligence

Fountain rebuilt its data plane on ClickPipes and ClickHouse Cloud, cutting analytics latency from three hours to under two minutes and reducing costs by 66%. This transformation powered their "Frontline Superintelligence" product, Cue.

Why it matters — Significant improvements in analytics latency (from hours to minutes) and cost reduction (66%) are achievable by migrating data planes to modern analytical databases, demonstrating clear ROI for such architectural shifts.

ClickHouse · 29 September 2026 · Read the original →

Reliability

Can your Postgres survive a bad query?

This benchmark compares how managed Postgres providers like RDS, Cloud SQL, and ClickHouse Managed Postgres handle recursive queries that exhaust memory. It evaluates whether the database engine can isolate and terminate a "bad" query or if the resource exhaustion leads to a full cluster outage.

Why it matters — Managed database "survival" under memory pressure varies significantly; some providers allow a single bad query to impact the entire cluster's availability.

ClickHouse · 29 September 2026 · Read the original →

Storage Internals

Inside branch-based restores in Lakebase Postgres

This post analyzes the architectural limitations of traditional point-in-time recovery (PITR), which becomes prohibitively slow as database volume grows. It explains how branch-based restores in systems like Lakebase Postgres use copy-on-write snapshots to decouple recovery time from data size.

Why it matters — For large production databases, traditional PITR is often too slow to meet RTO requirements, necessitating a move toward snapshot-based branching.

Neon · 29 September 2026 · Read the original →

PostgreSQL's intricacies continue to fill our weeks.

7 stories, every Tuesday

Published here every week. Follow by RSS to get it as it lands.

← Previous issue Next issue →