Feb 2021 Citus Newsletter (now part of Microsoft)

How to undistribute a distributed Postgres table

As one user asks, "what if you need to convert a Citus distributed table to not being distributed?" The new undistribute_table function makes it easy for you to experiment with distributing (and undistributing) your tables.

Read Ozan's first-ever blog post ⟶

Faster data migrations in Postgres

When migrating data from one Postgres database to another, these tips on speeding up offline data migrations are a must. The advice covers pg_dump, pg_restore, & the python Parallel Loader script.

Read Sai's post ⟶

Reconnecting your app after a failover

A biblical guide to understanding what happens when a Postgres failover occurs. Explains how to ensure your app reconnects with a read-write connection. Also, how to use pg_auto_failover to test your app's behavior after failover.

Read Dimitri's post ⟶

Migrating storage objects from Oracle to Postgres

More and more teams are migrating from Oracle to Postgres, which is why Arun created the Oracle to Postgres migration guide. However, there are things to watch out for in these migrations, including schema differences, numeric vs. BIGINT pitfalls, and NULLs in Postgres.

Read Arun's 2nd post in this series ⟶

Video of Marco's tech talk at Carnegie Mellon

CMU CS professor Andy Pavlo hosts a database seminar series called the "vaccination database talks" and last week's speaker was our Citus database lead. Marco Slot is one of the best explainers and if you want to understand how the Citus magic works, watch this video.

Watch Marco's talk ⟶

Async & direct IO support for Postgres

If you've ever wanted to peek into how Postgres engineering is done, this pgsql-hackers email from Andres Freund—announced via a late-night tweet, about his yearlong effort to figure out how to add async IO & direct IO support to Postgres—is well worth a read.

Read Andres's email ⟶

Postgres 13.2, 12.6, 11.11, 10.16, 9.6.21, & 9.5.25

The latest update to Postgres is out. This is the final release of Postgres 9.5 and everybody still running 9.5.x is strongly encouraged to update. This writeup gives you all the details about the 80 bug fixes & 2 security issues.

Read the news ⟶

Oldie but goodie on using rollups with Postgres & Citus

Learn how to build rollup tables using scalable incremental data aggregation, Citus, BRIN indexes—plus some Postgres magic. Includes benchmark results, too.

Read Marco's post ⟶

Tutorial for Citus on Azure

A new learning module for using Hyperscale (Citus) on the Azure Database for PostgreSQL managed service. Exercises include creating a Hyperscale (Citus) cluster, sharding and distributing tables, adding worker nodes, and rebalancing the cluster—without downtime of course.

Check out the learning module ⟶

Spring Data JPA tutorial with Azure Database for PostgreSQL

If you've been wanting to try out the Spring Data Java Persistence API with Postgres in the cloud, this tutorial should get you started.

Try the Spring Data JPA tutorial ⟶

Efficient rollup tables with HyperLogLog

HyperLogLog is an awesome approximation algorithm that addresses the distinct count problem. This post uses GitHub events data to show you how rollup tables and HLL can be used together for distinct counts.

Read Burak's post ⟶

ICYMI: New Oracle to PG migration guide for Azure

Oracle to Postgres migrations are a hot topic. To start, learn about the process from those who have done it—from pre-migration planning, to the migration itself, plus all the testing that comes after.

Read Arun's 1st post in this series ⟶