I planned on trying out Decodable to maybe some later day migrate a previously batched pipeline. It worked fine and was cheap, but I never had real-time data for testing.
After about 3 hours of tinkering on a slow Tuesday afternoon, not only did I finish my tests, but actually migrated the whole pipeline. And reduced technical debt in the meantime, as there now are fewer systems to maintain.
I checked in with my stream a few times, but never had to adjust anything (so far). It just works. Review collected by and hosted on G2.com.
My stream only processes about 1 to 2 records per second - even the small tasks are overpowered for that. To save energy and resources, even smaller task sizes (maybe by sharing task resources?) would be cool :) Review collected by and hosted on G2.com.
Thank you, Martin, for sharing your experience with Decodable! We're very happy to hear that your onboarding process was smooth. Our goal has always been to make stream processing as easy as batch processing, and feedback like yours reinforces our commitment to that goal.
Improving resource utilization is on top of mind. We’ve recently introduced a tech preview of Scale to Zero (https://docs.decodable.co/administer/scale-to-zero.html), which enables cost savings with the latency tradeoff. We’d love for you to try it out and share your thoughts!