June 2018 Newsletter from the database experts at Citus Data

Scalable incremental data aggregation on Postgres and Citus

This post hit the front page of Hacker News, for good reason. In it, Marco walks us through how to build rollup tables in a new and different way, to address the limitations of time windows. Why? So you can easily backfill older events and immediately update the rollup table when new events come in, providing a more real-time view of the data. Sweet!
Marco explains this new approach →

Fun with SQL: Functions in Postgres

There are several hundred handy functions built into Postgres. Craig walks us through a small sampling of SQL functions that can be extremely handy in PostgreSQL. So, next time you need to perform some operation on data—before you jump to writing a custom function, before you create custom logic in your app—see if Postgres already has your answer.
More on functions in Postgres →

Citus what is it good for? OLTP? OLAP? HTAP?

As Citans we often get asked, “Is Citus a good fit for X or Y or Z?” What folks want to know is whether the Citus distributed database is a better fit for analytical workloads or for more transactional workloads. Since we hear this question so often, Craig wrote a thing.
What is Citus good for? →

Marco Slot talks sharding on the Software Engineering Daily podcast

In this podcast, Marco shows Jeff Meyerson how Postgres—one of the most beloved relational databases—can be used to service almost all the needs of a data platform. This isn’t easy to do. It requires sharding your growing relational database into clusters and orchestrating distributed queries among those shards. Marco and Jeff explore Citus’s approach to the distributed systems problems of a sharded relational database.
Tune in to the SEDaily podcast →

Sign up for the live demo on Jul 25th: 10X faster analytics queries on Postgres

Join Samay Sharma, Lead Solutions Engineer at Citus Data, for a live demo of Citus Cloud, our fully-managed database as a service that scales out Postgres. This demo will compare the performance of analytics queries for a Citus database cluster vs. single-node Postgres. We’ll talk about how Citus shards Postgres—and how query performance scales linearly with the size of the Citus database cluster. And we’ll explore how Citus improves query performance by 10X or more.
Sign up today for the live demo →

Configuring memory for Postgres

At its surface, the work_mem setting seems simple: after all, work_mem just specifies the amount of memory available to be used by internal sort operations and hash tables before. And yet, leaving work_mem unconfigured can bring on a host of issues. This post gives good guidance on how to tune memory in Postgres, with a flyby on how Citus Cloud intelligently tunes the work_mem setting.
Craig’s post on memory configuration →

Fun with SQL: Window functions in Postgres

In this post in our ‘Fun with SQL’ series, we look at window functions in PostgreSQL. Key to analytic and reporting use cases where you want to compare and contrast data, window functions allow you to compare values between rows.
Read Craig’s post on window functions →

How do you pronounce Citus?

It’s a common question we get at conferences, on calls, in meetings. “Citrus”, “Citius”, “Citus”, is that how you pronounce it? The quick and short of it is, we’re not named after a fruit. You pronounce Citus like site-us. Inspired by Citius, Altius, Fortius.
But there’s more... →

ICYMI: Fun with SQL: recursive CTEs in Postgres

For applications that deal with e-commerce and shopping categories, recursive CTEs are a great help. Craig explains why next time you need to do some recursive computation across your data, you should consider doing it directly in SQL rather than loading all that data into your app.
Fun with SQL and recursive CTEs →

Videos, podcasts, and tweets. Oh my!

Is Citus right for you?

Want to explore whether Citus is right for you and your use case? Learn more about the Citus product, check out the documentation, or let us know if you want to book a demo.