Upgraded Insights, cheaper Macs, and more

Insights and Data

Understand performance patterns with Job Insights

Understanding your workflow performance is crucial for optimization. That’s why we’ve launched a new dashboard that allows you to dive deep into tens of thousands of past runs, aimed at helping you identify performance bottlenecks, optimize costs, and make data-driven decisions to improve your CI/CD pipeline.

Filter the list of jobs by one or more available dimensions and visualize your data in the graph through multiple views, broken down by your preferred options.

Insights are currently in Preview and available for everyone to try. Explore Insights for your GitHub jobs!

Deeper usage analysis

We’ve also added a few more options to the Usage explorer:

• View data with more granularity by comparing the current period to the previous one;
• You can now explore job counts, build counts, and other dimensions too;
• See the total compute usage for the selected period at a glance;
• Download the data as a CSV file to enrich it with your own data points.

Take a look at your Compute Usage.

Mac

Try out a Mac

Looking to quickly test something in a Mac environment? Create an ephemeral Mac instance with just one click in our web app.

And at just 37.5% of GitHub’s price, you get powerful caching and competitive pricing, so it’s never been a better time to speed up and reduce costs in one go — try it out!

Connect to a Mac instance right from the web

Easily connect to any Mac instance via Remote Display right from the web interface.

Do you prefer the CLI? Then simply run nsc vnc {instanceid}, and you’re in!

High-performance caching on Mac

The magic of cache volumes is now available on Mac, too. Attach a cache volume to your run, and get high-performance cross-invocation caching with little to no effort. Learn more about Cross-invocation Caching and how it can speed up your workflows.

It can be configured via GitHub Runner Profiles and is supported out of the box in nscloud-cache-action.

See it in action in this simple Rust demo.

Accelerating Bazel too

At Namespace, accelerating developer workflows is our bread and butter. We deploy the hardware, build the orchestration and infrastructure, and combine them into easy-to-use products that make your team go faster with little to no effort.

We started with Docker and GitHub Actions, following an early focus on Kubernetes, and extended to Buildkite and other custom workflows.

But we kept hearing from our customers – can you accelerate our Bazel too?

So we did. We recently extended our out-of-the-box support to include Bazel caching (as well as others, such as Buck and Pants, that use Bazel’s remote caching protocol).

We built Bazel support on top of a combination of cache volumes and our high-performance object storage cache. The two work together to provide a best-in-class experience that works right out of the box.

To start using Bazel caching, reach out to us at support@namespace.so. And let us know how you’re using Bazel today; we’re happy to help!

And keep an eye on our changelog; we’re working on Remote Execution, too.

Infrastructure improvements

We’re infrastructure geeks here at Namespace. Not everyone knows that we deploy our own hardware to data centers to support your workloads. We do that to ensure you run on the latest, most performant hardware and that your workloads have the right amount of network capacity to run as quickly as possible.

More network, fewer problems

We recently overhauled our network infrastructure, and each instance (or job) can now burst up to 10Gbps. You’ll see higher bandwidths coming to larger instance shapes soon.

Designed for Trust and Security

Our team has a long history of designing and operating large distributed systems, with a particular keen focus on security and privacy principles.

And those are the same principles we’ve applied to Namespace from day 1.

To uphold those, we get external reviews and security tests done on a regular basis, and you can find the results in our SOC2 Report in our Trust center.