---
title: Exploring Postgres VACUUM with the VACUUM Simulator
---

| Learn how to ensure updates are HOT capable, our latest eBook on autovacuum, celebrating 100 episodes of 5mins of Postgres and more! |
| --- |

| View in browser Hi there, Welcome to the January 2024 edition of the pganalyze newsletter!   Today, we're happy to announce a new eBook about tuning vacuum for best Postgres performance, are happy to celebrate 100 episodes of 5mins of Postgres, and have a new pganalyze feature to optimize your workload and index choices for HOT updates (see more details below).   On the company side, we're happy to welcome our newest team member, Nico, who joined us in January to lead sales. If you want to talk to us about purchasing pganalyze, or are interested in a demo, you can reach out through our Contact Us page, or reply back to this email. And, if compliance is on your mind: Our SOC 2 Type 2 report for 2023 is now available!   As always, you will also find other product updates, and all recent "5mins of Postgres" episodes in this newsletter. All the best, Lukas New in pganalyze: Built-in assessment whether updates are HOT capable We've launched a new feature in pganalyze to show you when updates on a table's column are HOT capable.   Not sure what that means? Heap-Only Tuples, aka HOT, are an important performance optimization in Postgres, that makes updates much more efficient. However, they are only possible if the columns being updated are not indexed (= the update is "HOT capable"). If you want to do a deep dive, we talked more about the benefits of HOT updates in Episodes 72, 40 and 14 of 5mins of Postgres.   To help you optimize your workload and index choices for HOT updates, you can now quickly see which columns are indexed on the "Columns" page for each table. This also considers columns indirectly referenced through an expression, a partial index, or in the INCLUDE part of a covering index. On the columns page you can also see how many updates each column receives per minute, and whether these updates are HOT capable.   This assessment is also done for each index recommendation considered by pganalyze Index Advisor as part of the new constraint programming model - more news on that in next month's newsletter! Read the documentation 📚 Free eBook: Tuning Autovacuum for best Postgres performance You might remember the launch of our pganalyze VACUUM Advisor back in July 2023. A lot of research has gone into developing this feature and in this new eBook we are sharing our learnings around the most important concepts of why and when Postgres has to vacuum, why that matters, and when it’s best to tune the default settings that control autovacuum scheduling, vacuum overhead, and more. In the book, we provide actionable tuning tips you can use in your specific setup. Download the free eBook Celebrating 100 episodes of 5mins of Postgres When we initially started publishing our 5mins of Postgres videos, our goal was to share a quick, bite-sized video about recent Postgres topics once a week with the wider community.   We've been surprised and are excited about how well this weekly series has been received! Today, we have over 3,000 subscribers on our YouTube channel and are happy to be interacting with so many of you directly there. In case you missed it, we also started uploading our webinar recordings to this channel - so feel free to subscribe and check out our videos!   Looking back, some of our favorites among the first 100 episodes are:   E82: Postgres 16 is released!→   E74: How Postgres has adapted to AI and ML use cases using pgvector→   E59: UUIDs vs Serial for Primary Keys - what's the right choice?→ 5mins of Postgres E98: How to partition Postgres tables by timestamp based UUIDs→   We talk about partitioning Postgres tables by timestamp based UUIDs. We're also going to talk about the status of UUIDv7 in the current Postgres development branch. E99: Finding the root cause of locking problems in Postgres→   We show how to go about finding the source of a locking problem in Postgres. Specifically, we are looking at how to debug heavyweight locks, how to end a process that’s holding a lock in Postgres, and general helpful settings for handling Postgres locking. E100: Reducing table size with optimal column ordering and array-based storage→   We take a look at how column ordering affects storage size, how you can play column tetris to optimize it, and the big benefits that array-based storage can yield for some workloads. Subscribe on YouTube We publish weekly videos on Postgres, our conference talks, tutorials, product walk-throughs, and more. Subscribe to pganalyze on YouTube Product updates Schema Statistics New HOT update information on Columns page Speed up loading of per-table list of scans when many scans are present pganalyze-collector release 0.53.0 Track extended statistics created with CREATE STATISTICS This is utilized by pganalyze Index Advisor to better detect functional dependencies, and improve multi-column index recommendations To allow the collector to access extended statistics data you need to create the new "get_relation_stats_ext" helper function Docker image: Don't reload when calling "test" command Other improvements and bugfixes Alerts & Check-Up: Fix layout bug in organization-wide check config listing VACUUM Advisor: Change sorting on activity page to list most recent vacuums at the top of the list EXPLAIN Plans: Improve visualization Padding between nodes is reduced so more content fits on screen Table and index references now link to their respective pages CTE scans now link to their parent node, making it easier to navigate large plans Review us on G2 Please leave your feedback (good or bad!) in this survey - we appreciate it! Fill out the G2 survey Duboce Labs, Inc., 1501 Mariposa St, Unit 408, San Francisco, CA 94107 Unsubscribe Manage preferences |
| --- |