Blogs by: Frits Hoogland

Client-Side vs Server-Side Latencies Demystified in PostgreSQL and YugabyteDB

Client-Side vs Server-Side Latencies Demystified in PostgreSQL and YugabyteDB

Every SQL execution in PostgreSQL and therefore in YugabyteDB YSQL takes time to process. A common way to identify how much is time spent on processing is to use the pg_stat_statements view in the database. However, the time visible in pg_stat_statements might differ from the time a database client registers for the execution. Where does this difference come from? Let’s take a look.

Read more

Performing Flexible IO Testing in YugabyteDB

Performing Flexible IO Testing in YugabyteDB

ybio is a PL/pgSQL based load generator for PostgreSQL and YSQL. This YugabyteDB-specific IO testing toolkit performs flexible IO testing to specific parts of the YugabyteDB infrastructure with no client installation. And because this toolkit is PL/pgSQL-based, it can also work on native PostgreSQL.

ybio is strongly inspired by Kevin Closson’s SLOB and pgio. Yugabyte Developer Advocate Franck Pachot took pgio and turned it into ybio.

Read more

Database Connection Management: Exploring Pools and Performance

Database Connection Management: Exploring Pools and Performance

There are many secrets to creating a high-performing database application. One such secret is proper database connection management. However, this secret is not specific to YugabyteDB or PostgreSQL, but applies to any database.

In this post, we examine database connection management through the lens of YugabyteDB. We explore pools and performance, and run tests to measure the results. By the end of this post, you’ll have a clear sense of how to make efficient use of database connection management.

Read more

Linux Performance Tuning: Dealing with Memory and Disk IO

Linux Performance Tuning: Dealing with Memory and Disk IO

Linux is a general purpose operating system. This means it’s created to generally do what is right, instead of having specific code paths to perform what is right for a single specific task—and potentially be wrong for others.

Linux does not have tunable parameters for reserving memory for caching disk pages (the page cache), like operating systems such as HPUX (dbc_min_pct, dbc_max_pct) or AIX (minperm%, maxperm%). Instead,

Read more

The YugabyteDB Vagrant Box

The YugabyteDB Vagrant Box

Using Vagrant, anyone running Windows, Linux or Mac OS X can run and test drive a YugabyteDB cluster without the need to install and configure YugabyteDB. Plus, you can start over if you want to test something different in a matter of minutes, or scale up the cluster, etc.

Please mind that the Vagrant images are non-official images for demo and test purposes, and not configured nor representative for performance.

This is how it works (Mac OS X):

1.

Read more

Explore Distributed SQL and YugabyteDB in Depth

Discover the future of data management.
Learn at Yugabyte University
Get Started
Browse Yugabyte Docs
Explore docs
PostgreSQL For Cloud Native World
Read for Free