Improvements
Today we had significant downtime (57 minutes) for the first time in years, something we consider unacceptable. To prevent this from happening again, we have made several changes to our infrastructure and processes:
Instead of a simple load-balanced setup for our Ingress API, we have moved to a queue-based system handled by Cloudflare Workers, with fallback to AWS Lambda, and with final failover back to the original endpoint. This makes the system more resilient to spikes in traffic and failures in individual components.
This means that the Ingress API is no longer a single point of failure, and can handle much higher loads.
You should also see improved latency when uploading coverage data, due to the edge-based nature of Cloudflare Workers.
This process requires no changes on your end, and your existing upload scripts will continue to work as before. We will continue to monitor this new setup closely.
We have also added a link to our Uptime Statuspage in the footer of our website, so you can always check the current status of our services.
In short, this means our database, website, API, dashboard and other components can crash or be taken down for maintenance, without affecting your ability to upload coverage data.
We regularly update our changelog in an effort to improve transparency into the development process. If you have any questions or feedback, please don't hesitate to contact us.
If you are a user of OtterWise already, or have been reading our blog, you might already know that we take privacy extremely serious, even when it comes to protecting your code as if it was sensitive personal information, this goes for everything we do, which is why we publish changelogs for even smaller updates, rather than just major releases.
Our weekly digest for OtterWise users also includes our changelog in the bottom (see related changelog), to keep you up-to-date with the latest changes.
