This is a short story of how Dashbird was born and what we promise in the future.
The problem
When we first started dealing with AWS Lambda functions we quickly realised that whereas the technology was awesome, the tooling was not.
After reaching a certain scale it just became increasingly difficult and borderline impossible to debug our serverless services. We just had too many lambda invocations to use cloudwatch. Since cloudwatch was not working out for us, we decided to look around for other service providers to help us debug our services.
We tried all the services on the market to help us on that quest and most of them provided beautiful graphs, but that’s also where the offering would end. We also did not fancy the idea of changing our now legacy lambda functions to include third party libraries to them.
The promise
We decided that there must be a better way for debugging and thus Dashbird was born. The main principles of Dashbird were and always will be:
- Simplicity — our goal is to be the easiest monitoring and debugging system out there. This means that we do not like complicated setup steps nor do we like code changes for monitoring tools.
- Visibility — we aim to highlight all invocations, errors and mistakes explicitly. This means allowing users to search and browse all invocations on their own.
The outcome
We launched Dashbird to private beta in the beginning of March 2017. During the last five months we have spent countless hours improving our service for ourselves and our users.
Now, five months later, it is an absolute pleasure to publicly announce that Dashbird.io is finally going out of beta. This means that we now have more time and resources to build the best monitoring system for serverless solutions out there.
We would like to express our greatest gratitude to all of our beta users — you are awesome. Every single one of you! The feedback that our beta users gave us was valuable in every way.
The future
Dashbird is evolving daily, but there are a few huge features rolling out within the next few weeks.
- Alerts — we believe that having a monitoring service without alerts is unacceptable. Therefore we will be releasing our alerting solution soon.
- JAVA and C# support for AWS Lambdas — currently we support python2.7, python3.6, node4.3 and node6.1 but we intend to support all AWS Lambda runtime environments out there.
- Project views — having hundreds of lambdas is cool, but whats more awesome is to have a way to quickly see what different stacks are doing and to focus only on those that matter
- Multiple accounts — currently its impossible to quick switch between clients. This will change soon. Within the next few weeks it will be possible to quick switch between all different accounts where a user is a member of.
- Azure Functions—we were not kidding when we told that we want to become the best serverless monitoring platform out there. Becoming the best also means handling all the different platforms that serverless supports, so we are starting with Azure.
Of course we intend to improve our service in terms of speed and usability meanwhile. This means having lower latency between imports and quicker response times from our servers.
You can relate?
Programming is basically a process of creating bugs. So naturally, there is a need for a good debugging tool. Dashbird was born out of real need and our own frustration. This means we promise you no fancy bells and whistles, but just an honest and simple debugging possibility.
