PostgreSQL 19 dev: Have the planner replace COUNT(ANY) with COUNT(*), when possible Author: @davidrowley_pg https://t.co/P5ko38aXWD

It was a pleasure to give a talk last week at @PosetteConf about what's new in the PostgreSQL 18 query planner. There are some great new features in this release.

🐘 🤔 If you're interested in the #Postgres Query Planner, you'll love David Rowley's (@davidrowley_pg) #PosetteConf talk. Discover why David chose to talk about the Query Planner and more at: https://t.co/mscIDEsn4b #postgresql #databases #OpenSource

If you're a #PostgreSQL developer, community member, or curious user—you'll likely enjoy the #TalkingPostgres #podcast where we talk to people who work with Postgres—to learn why they do what they do & how they got there. Subscribe today so you'll never miss an episode!

I had no idea until today that such a large part of the world didn't class Monday as the first day of the week. ISO8601 and RFC3339 say it's Monday. I always thought the term "weekend" was fitting for days 6 and 7, but that doesn't seem to fit as well in blue or green areas.

There are a few PostgreSQL functions that could be suboptimal because of this: btint4cmp(), btint8cmp(), bms_compare(), perhaps more.

Clang 18.1 does the same thing (effectively), but clang trunk produces the same code. (I added the # comments to the screenshot as per my understanding of what's happening.)

Using https://t.co/63BPwdER0X to see how well a typical comparison function to sort integers in C is optimised. I was surprised to see the order I wrote this affected the produced assembly. You can see cmp1 and cmp3 require 2 CMPs because NEG overwrites status flags. Interesting

The 2nd CommitFest of Postgres 18 starts today. I noticed an increase in performance-related patches. The chart below shows the top 5 topics from the last decade of CFs. We've had ~40 performance patches committed per year < 2024. This year we've had 64 already with 2 CFs to go!

One of the things I'm working on for PostgreSQL 18 is executor performance. Today I pushed a patch to add JIT support to increase the performance of generating hash values for Hash Join. Thanks to @AndresFreundTec for letting me know this needed attention. https://t.co/CEyKebKC25

Thanks to @clairegiordano and the Talking Postgres team for giving me a chance to tell my story about how I ended up working full-time on PostgreSQL. There certainly wasn't a single defining moment, but likely a few key things that took me in that direction.

In just over a week, I'm getting a rare opportunity to tell my story about how I ended up working full-time on PostgreSQL. Many thanks to @clairegiordano and all the Talking Postgres team for making this happen. I'm looking forward to it.

Recent blog post from @davidrowley_pg from the Postgres team at Microsoft, which developers & users loved 💟 All about what's new 🤯 in the #PostgreSQL 16 query planner As @SQLPerfTips says, it's "short, relevant, clear. A MUST read for everybody"

Lukas of @pganalyze gives a great summary (in video form) of a blog I published yesterday about what's improved in the PostgreSQL 16 query planner.

This is one of these very few articles: short, relevant, clear. A MUST read for everybody—whether using PostgreSQL or not. It is generally good to see how these things work. Thank you @davidrowley_pg — for the patches too, of course! https://t.co/HbbcQPKc3H

This deep dive by @davidrowley_pg on PG16 improvements is great! https://t.co/pNvAs30YDi What I love just as much as him showing up in HN comments, not defending or deflecting just being there and taking input but also educating

In @davidrowley_pg’s new blog post about planner improvements in #postgres 16, he provides example queries and plans for every optimization mentioned. It’s a great way to learn about planner if you’re new or to brush up on minimal repros if you’re seasoned

If you're thinking of upgrading to PostgreSQL 16, I wrote this blog to help explain (with examples) what's changed since PG15 in the query planner/optimizer

David Rowley (@citusdata): What’s new in the Postgres 16 query planner / optimizer

New post from @davidrowley_pg from the #PostgreSQL team at Microsoft, about what's new in the Postgres 16 query planner (aka optimizer) tl;dr ✅ PG16 has quite a few performance improvements to the query planner, this post walks through 10 of them