Better Networks Through Network Observability
Networks are a fundamental component of modern infrastructure, in businesses, to individuals, and society as a whole. But despite their importance, networks are brittle. The way they’re designed, built and operated, makes them brittle. While there are many reasons for this, one of the fundamental reasons for this is the lack of high quality tools to help build, design and operate networks.
A lot of money is being spent on software development productivity and several unicorns are based on that such as Jfrog and Hashicorp. The same needs to be done for networking. Networking has its own challenges, so you cannot just move straight tools or processes from software development, but there is much we can learn. And more importantly, the productivity of Network Engineers is too low. Because it’s so low, networks are harder to operate, harder to change, and harder to run well.
Operator error is the #1 or #2 reason for network failures, ahead of hardware failures or software bugs. Operators lack the tools to provide the guard rails or the ability to address the source of these errors. Many networks still involve a lot of manual operation. This introduces random errors which are hard to catch. When network operators turn to automation, operator error can result in widespread failures if errors are not caught early enough. Sadly, network operators lack the tools to use network automation to deploy configuration with confidence. When it comes to changes, the problems just get worse. If you don’t have a way to know that your network is working as expected, when there’s an error after making a change, it’s hard to know if the change caused the error or did it always exist but get exposed because of the change.
And when it comes to designing networks, whiteboard and spreadsheets are the most common tools, shaped by the vendor they’re locked in with. Designing requires them to first push back against vendor-specific solutions which tend to play to a vendor’s strength rather than the needs of the customer. If network designers had tools that allowed them to compare and contrast vendor choices in a neutral way, enterprises can better build networks appropriate to their needs.
Networks need better. Network engineers and operators need much better software. Stardust makes software which gives network engineers super powers, so that we get the networks we need.
Suzieq for Network Observability
We’re building Suzieq which is open source software focused on helping you understand how your network is actually operating
Join us and help us make it great.