© Laurenz Albe 2023 “Why does my pg_wal keep growing?” That’s a question I keep hearing again and again. It is an urgent question, since the PostgreSQL server will crash if pg_wal runs out of disk space. I decided to answer the question once and for all. What is pg_wal and why is it growing? […]

© Laurenz Albe 2022 Most people are aware of transaction ID wraparound. The concept has been well explained in Hans’ article, so I won’t repeat all that here. But for most people it is an abstract concept, a bogeyman lurking in the dark around the corner. Many people know horror stories about anti-wraparound autovacuum tanking […]

If you are running PostgreSQL in production you might have already noticed that adjusting checkpoints can be a very beneficial thing to tune the server and to improve database performance overall. However, there is more to this: Increasing checkpoint distances can also help to actually reduce the amount of WAL that is created in the […]

© Laurenz Albe 2019   “Durability”, the D of ACID, demands that a committed database transaction remains committed, no matter what. For normal outages like a power failure, this is guaranteed by the transaction log (WAL). However, if we want to guarantee durability even in the face of more catastrophic outages that destroy the WAL, […]

UPDATED September 2023: Most people who use PostgreSQL database systems are aware of the fact that the database engine must send changes to the so-called “Write Ahead Log” (= WAL) to ensure that in case of a crash, the database will be able to recover to a consistent state safely and reliably. However, not everybody […]

To use PostgreSQL replication it is necessary to change the wal_level from “minimal” to “hot_standby”. But, which impact does this change have on the amount of WAL actually written to disk? And which impacts does it have in performance? Many people have been asking this question in the past, so I thought I’d give it […]