Nov 2021 Citus Open Source Newsletter

How Citus Distributes PostgreSQL

In this video, inspired by his SIGMOD paper from earlier this year, Marco Slot explains how Citus distributes Postgres:

00:30 What is Citus?
01:17 Four workload patterns that need to scale PostgreSQL
05:42 How Citus works
14:28 Performance benchmarking of Citus

Introducing Private Link in Hyperscale (Citus) on Azure, a “how-to” post

With the preview of the Azure Private Link for Hyperscale (Citus), you can now bring your Hyperscale (Citus) server groups—new or existing—into your private Virtual Network space. In this step-by-step blog, Rongxun explains how you can create and manage private endpoints for your Hyperscale (Citus) database nodes.

Read Rongxun’s post ⟶

pg_auto_failover 1.6.3 just released—with this intro to the project

pg_auto_failover 1.6.3 was released earlier this month on GitHub. You can find all the bug fixes and enhancements here. In this article, Dimitri introduces the project, and answers the why, when, what, who, and how for the project overall.

Read Dimitri’s post ⟶

Postgres customer “cascade of doom” post (with a happy ending)

The UK Health Security Agency, a Hyperscale (Citus) customer, had a scary Halloween. After upgrading to a new version of Postgres they were unable to deploy new data to the UK coronavirus dashboard. Here’s how their technical lead and our Postgres team diagnosed and fixed the problem, and restored a critical national service—and what you can learn from their experience.

Read Pouria’s post ⟶

Tweet from Marco: only open source features in Citus for ~2 years

For ~2 years, we haven’t added any non-OSS features to Citus. (You can follow along with ongoing development on GitHub.) Hyperscale (Citus) on Azure also has all the usual managed database services, and many of the tools we use are open source too.

Read Marco’s tweet ⟶

Our sister team’s Flexible Server GA blog post

Flexible Server for Azure Database for PostgreSQL GA’d today, Nov 30, 2021. Now available in over 30 regions around the globe. This is big news for Postgres users on Azure whose applications don’t need the parallelization and distributed scale of Citus.

Read Sunil’s post ⟶

Postgres end-of-year gift idea: A Curious Moon by Rob Conery

A Curious Moon is about data and Postgres. Cassini and Enceladus. You’ll learn Postgres the way the pros do: on the job and under pressure, using data from Cassini about Enceladus. Watch the video embedded in the site. It’s fascinating and inspiring.

Order A Curious Moon ⟶

Recent question on Reddit: “Do you know good articles on geospatial time series or spatiotemporal data?

i.e. asset tracking, wildlife tracking, traffic modeling? Would love an in-depth article on @postgis with Citus.” Our answer: “Analyzing GPS trajectories at scale with Postgres, MobilityDB, & Citus”

Read Mahmoud’s post ⟶

Let’s create a distributed columnar database using Citus

Earlier this month, at PGCon21J, Noriyoshi Shinoda from Hewlett Packard introduced how to create distributed columnar tables with Citus. Her slides, in Japanese, walk you through the process step-by-step.

Read Noriyoshi-san’s slides ⟶

ICYMI: Using partitioning & sharding together in Postgres for time series data

This post shows you how to manage your time series data in a scalable way. By combining the partitioning capabilities of Postgres, the distributed database features of Citus, and the automation of pg_cron—you get a distributed relational time series database.

Read Burak’s post ⟶

ICYMI: Postgres 14 now available for Hyperscale (Citus) on Azure

You can now deploy Postgres 14 in many Hyperscale (Citus) regions, with more being added all the time. So you have access to the latest Postgres features like improved connection scaling and faster crash recovery, on Azure.

Read Ozgun’s post for more deets ⟶