Cockroach Labs Blog

Cockroach Labs is hiring people to work on CockroachDB from Sydney, Australia! If you're curious about me -- the lone Aussie -- my journey, and why I'm at Cockroach, read on.

This post looks at three solution paths for orchestrating stateful apps in Kubernetes, and walks through some of the factors that might help a dev team choose between them.

As long as you're willing to follow the rules, deploying on Kubernetes and traveling by plane can be quite pleasant. But if you want to fly with your alligator or scale a database that must remain available, Kubernetes gets a bit more complicated.

We hosted several interns over the year across various engineering teams, all working on projects deserving of full-length blog posts. Today, however, we'll highlight two projects from our most recent batch and give a briefer treatment for the remaining.

CockroachDB supports some of the most data-sensitive industries including healthcare and finance. CockroachDB protects sensitive data by redacting it data out of log files and crash reports.

Now in its third year, the 2021 Cloud Report compares AWS, Azure, and Google (GCP) to tell realistic performance stories on behalf of mission-critical OLTP applications.

With our team working from home these past months, we’ve had to think creatively on how to foster learning opportunities on our Engineering team, specifically: when you first join the company and day-to-day learning from each other.

Support for spatial data and geospatial indexing was one of the most requested features in the history of CockroachDB. As of CockroachDB version 20.2, we launched support for scalable spatial data and indexing.

CockroachDB is PostgreSQL compatible in order to be accessible to an ecosystem of drivers, existing code, and developer knowledge.

Throughout our journey, our open source community has made valuable contributions to CockroachDB. This year, we had over 180 commits from over 65 contributors. Thank you.

I’m a database nerd. Or, to be more precise, a DBMS nerd. But back in university, I avoided the Databases course at all costs.

Distributed BACKUPs and RESTORE are now in CockroachDB Core to provide reliable, valuable Disaster Recovery for all CockroachDB applications.

Understanding the Raft Consensus Algorithm can be tough. In honor of Mean Girls day (October 3), we present Raft as explained through the movie Mean Girls.

The CockroachDB dialect for Hibernate is officially available 🥳 Hibernate now offers first-class support for CockroachDB, so get building in Java and CRDB.

Today we’re introducing Pebble, a RocksDB-inspired and RocksDB compatible key-value store written in Go that brings greater stability to CockroachDB.

Database schema design is difficult. Requirements evolve, and somtimes columns have to change types. CockroachDB enables ALTER COLUMN TYPE without taking tables offline.

Lateral joins - similar to nested foreign loops - are a type of correlated subquery that allow data sources to reference columns provided by previous data sources in the same query. Give them a spin in CockroachDB.

What do nested transactions look like in an ACID environment? We explore the journey of a nested SQL transaction in CockroachDB in this post.

"CockroachDB: The Resilient Geo-Distributed SQL Database," a research paper codifying years of research on CockroachDB, will be shared at ACM SIGMOD 2020.

As of our 20.1 release, CockroachDB supports online primary key changes. This means that it's now possible to change the primary key of a table while performing read and write operations on the table. In this blog, we'll review how data is stored in CockroachDB and how we overcame technical challenges involved in getting online schema changes to work in CockroachDB.

We want to give you a database that allows you to quickly and simply build and manage data-intensive applications in the cloud—without worrying about the performance, scale or resilience of those apps. To support that mission, our latest release includes updates that let you build fast and build to last.

If someone knows even a little about Spanner, the first question they have is: "How does CockroachDB provide external consistency without atomic clocks?"

How we built native support for hash sharded indexes in CockroachDB, which can significantly improve performance on sequential workloads.

Key learnings about how to get the best performance from GCP, AWS, and Azure based on original research and conversations with the big three.

At Cockroach Labs, we’re focusing on diversity and inclusion in our college and university recruiting. Here’s what we've learned so far.

The 2020 Cloud Report compares AWS, GCP, and Azure on microbenchmarks and customer-like-workloads to understand performance tradeoffs within each cloud.

Peewee ORM now supports CockroachDB. And now it's easy to get started with the distributed, horizontally-scalable SQL database.

In this blog post, we introduce Follower reads, a key feature for supporting multi-region reads with low latency when your use case can accept stale data.

This post dives into consensus algorithms and how we have added support for atomic replication changes (“Joint Quorums”) to etcd/raft, using them in CockroachDB to improve resilience against region failures.

Yugabyte frequently compares themselves to CockroachDB. We investigated their claims. This is our analysis of CockroachDB vs. Yugabyte v2.0.0.

SQL Prober: How we used geo-partitioning to build a monitoring system that fits our distributed architecture.

There are a few basic rules to keep in mind when choosing indexes for a database. A good index should have these three properties: Usefulness, Clustering, Scattering.

When the CockroachDB SRE team began building CockroachCloud, they evaluated different automation and orchestration technologies -- including taking a hard look at Kubernetes. They didn’t break any laptops over their knees, but it was quite the journey nonetheless.

In this post, we explore the steps we've taken so far--and plan to take--to improve the CockroachDB cost-based optimizer through query caching.

How a vectorized (column by column instead of row by row) approach to merge joins led to a 3x improvement in standard merge joins in CockroachDB.

CockroachDB is an excellent system of record, but no technology exists in a vacuum. That's why we built our version of Change Data Capture, which we call the changefeed.

Golang executable files get too large for reasons that are not relevant to server software with long-running processes, like CockroachDB, and its incurred space cost is particularly inconvenient for large, feature-rich programs.

After two years of nightly testing, we’ve come to learn the power and limitations of Jepsen as a testing tool.

How we built a 40x faster hash joiner using vectorized execution.

CockroachDB's Consistency Model fits somewhere between serializability and linerarizability. We're proposing a new marketing phrase for CRDB's guarantees: no stale reads.

CockroachDB uses RocksDB for its storage engine because of RocksDB's rich feature set, which is necessary for a complex product like a distributed SQL database.

This post introduces Transactional Pipelining which dramatically speeds up distributed transactions with respect to the latency cost of distributed consensus.

This blog covers the practical experience of running a distributed system across multiple Kubernetes clusters including what makes it challenging and what solutions are available (some of which we run in production).

CockroachDB 2.1 includes a brand-new, built-from-scratch, cost-based SQL optimizer. This post explains what a cost-based SQL optimizer is, and tells the story of how we decided we really, really needed one.

sqlfmt is an opinionated SQL formatter. It’s not designed to be completely customizable, and we hope to remove options over time. Use it and stop thinking about how to format SQL.

CockroachDB was so named because of its ability to survive failures. In this post, we will cover architectural considerations when building high availability services on top of CockroachDB.

If your default mode of thinking is SNAPSHOT, write skew anomalies are just normal behaviour and SERIALIZABLE is throwing away perfectly good histories.

Managing state in Kubernetes is difficult because the system’s dynamism is too chaotic for most databases to handle; especially SQL databases that offer strong consistency.

Learn which performance metrics matter, why to demand benchmark TPC-C results under serializable isolation, and how CockroachDB 2.0’s performance compares to 1.1.

With JSON support in CockroachDB, you can now use both structured and semi-structured data within the same database.

Scaling a database beyond one zone is a challenge. Here's some strategies on how to approach it.

Baidu and Cockroach Labs jointly held a conference that would bring top-tier DBAs and developers together to discuss CockroachDB.

Users with an existing database may have trouble getting their data into CockroachDB, so we built a new feature that performs high-speed, bulk data import.

This post is the first in a series on how CockroachDB handles its data and discusses the mechanisms it uses to rebalance and repair.

CockroachDB makes it easier to manage and scale mission-critical data, and DC/OS provides elastic scale, zero downtime, and hassle-free operations.

Most databases treat strong transaction isolation as exotic, expecting applications to work with faster but unsafe isolation modes that can expose them to subtle bugs. CockroachDB provides SERIALIZABLE transactions by default, ensuring you always see the consistency you expect.

This post explains how CAP-Consistent systems can be highly available, how the CAP theorem applies to CockroachDB, and why consistency is the better choice.

ActiveRecord is the ORM that powers Ruby on Rails, one of the most popular web app frameworks. What went into supporting ActiveRecord in CockroachDB?

CockroachDB 1.0 was built to be production-ready with an OLTP use case in mind.

The apd package replaces the underlying implementation of the DECIMAL type in CockroachDB and is available for anyone to fork and use.

As we prepare for CockroachDB 1.0, we hired Kyle Kingsbury, the author of Jepsen, to review our tests and add more of his own. Last week, Kyle published his results.

In this post we share performance and functionality progress that has been made on SQL joins in CockroachDB and what to expect in CockroachDB 1.0.

In this post, Joram Barrez shows how easy it is to work with Flowable and CockroachDB.

In August, we wrote about our concerted stability effort. Today, we've more than achieved our CockroachDB stability goal of running a 100-node cluster.

Some months ago I started work on a way to test random, valid SQL in CockroachDB. This is important to expose unintended behavior in our server.

Our approach to building CockroachDB has been to focus on correctness, then performance, then stability. Without stability, we don’t have a working database

The good news is that CockroachDB’s JOIN seems to work, as in, “it returns correct results.” However, this is just our first, unoptimimized implementation.

If we abandoned our earthly constraints, how could we improve consensus algorithms? Specifically, what would it take to make them faster?

This blog post explains how to run CockroachDB in a FreeBSD jail. This is lighter weight and as secure as running Docker on FreeBSD.

Writing good tests is tricky when the system has a lot of moving parts. In this post we will go over how to outsmart Go dependencies in testing.

Adopting SQL had an unexpected consequence; it forced us to dabble in language design. In this post, we detail our approach to SQL typing in CockroachDB.

CockroachDB's support for SQL means access to a variety of data frameworks and tools. Today, we'll build a simple application in Python using SQLAlchemy...

This week, Mesosphere Developer and Cloud Advocate Michael Hausenblas made it his mission to run CockroachDB on DC/OS.

CockroachDB is pretty easy to deploy. However, there is still one wrinkle in the fabric, and that's our use of network ports.

To protect concurrent transactions from anomalies that can break the consistency of CockroachDB, we require an Isolation guarantee...

In this post, we will go over some of the factors involved in choosing the best index to use for running a certain query.

We at Cockroach Labs absolutely love Aphyr's work. We are avid readers of the Jepsen series – which some know as high quality reviews of correctness...

The digest provides additional insight into the focus of a pull request by listing the most important subdirectories. File changes are tallied for each subdirectory, and those which comprise the top 80th percentile are deemed representative of the pull request and listed immediately underneath the additions and deletions totals

This is the story of the detection, investigation, and fix of just one performance bug.

GitHub stars are a simple way for GitHub users to bookmark repos that interest them. "Stargazers" is a tool to query GitHub for information about the stars.

I experimented with implementing the Redis protocol on top of CockroachDB, scratching the itch on the question: Could CockroachDB ever replace Redis?

Cgo is an important part of Go: It's your window to calling anything that anything that has C bindings. But while cgo works for us, it hasn’t come for free.

In this post, we’ll share a few powerful optimizations that mitigate many of the performance problems common to Go garbage collection.

The first question many developers ask us is what our experience has been writing a distributed database in Go, a garbage-collected language.

SQL? I thought CockroachDB was a key-value store?!? In the past we described CockroachDB as a distributed, transactionally consistent, key-value store.

One of the headline features of CockroachDB is its full support for ACID transactions across a distributed database. In this post, we’ll explain...

In CockroachDB, we use the Raft consensus algorithm to ensure that your data remains consistent even when machines fail.

Today, we’re launching CockroachDB for everyone. Use it. Build on it. Contribute to it!