Postgres Replication Slots: Cap WAL Retention Before Disk Full
Replication slots retain WAL unconditionally. Size max_slot_wal_keep_size from peak WAL rate, alert on wal_status, and cap retention before the primary fills.
Expert PostgreSQL tutorials, tips, and performance optimization guides. Learn database administration, query optimization, and best practices for PostgreSQL 18 and beyond.
High‑availability setups, backups and recovery, major/minor upgrade strategies, and Kubernetes/containers content.
Patterns for multi‑tenant SaaS, read scaling, partitioning, and observability/monitoring.
Replication slots retain WAL unconditionally. Size max_slot_wal_keep_size from peak WAL rate, alert on wal_status, and cap retention before the primary fills.
pg_restore exiting 0 is not the finish line. A Postgres restore needs statistics, a visibility map, integrity checks and cutover gates before traffic — here is the runbook, with measured numbers.
A default GUC on nearly every PostgreSQL standby can turn a brief network glitch into a 30-minute write outage. How recovery_prefetch amplifies replica IOPS, and how to decide whether to keep it.
Release candidate 1 is out and GA lands November 19. A five-week ops plan to rehearse pg_upgrade, replication cutover and rollback before production sees it.
pg_vault_tde 1.7.2 shipped a new on-disk format and a WAL resource-manager ID change that rules out a rolling upgrade. An operations runbook for Postgres transparent data encryption: config, upgrade steps, key rotation, backups and monitoring.
Seven CVEs, six of them in watchdog: the Pgpool-II subsystem that owns quorum, the delegate IP, and backend promotion. How to patch a live watchdog cluster one node at a time and harden the watchdog plane while you are in there.
Most failover drills end the same way: the standby takes over, applications reconnect, and everyone goes home. The part that quietly broke is backup continuity. A community question summarized this week by Stefan Fercot on his pgBackRest blog describes exactly this: a standby was promoted with archive_mode=off, and the team wanted to avoid restarting the […]
PostgreSQL 19 slipped, so plan the upgrade you can actually control: a logical replication cutover with sizing, rehearsal, runbook and rollback.
A three-node Patroni cluster on commodity VMs costs less than one hour of downtime. Here is the architecture, the config, and how to prove it works before you need it. Tom Lane recently walked through the architectural decisions that shaped 30 years of Postgres, and one theme runs through it: Postgres was built to be […]
On August 13, 2026 the PostgreSQL Global Development Group released 18.6, 17.11, 16.15, 15.19, 14.24 and 19 Beta 3. If you run logical replication, manage pg_upgrade plans, or operate in FIPS-enforced environments, this is not a routine minor release — two of the security fixes change runtime behavior in ways that can break production after […]