Inpreviousblogposts,we discussed how SoundCloud has been moving towards a microservice architecture.Soon we had hundreds of services, with many thousand instances running andchanging at the same time. With our existing monitoring set-up, mostly based onStatsD and Graphite, we ran into a number of serious limitations. What wereally needed was a system with the following features:

The article How to be a More Reliable Professional, says that 'Unreliability was identified as the number one career-limiting habit,' which I found very surprising, but also very understandable.No one wants to. The night before. I am a ball mixed with excitement, smiles, jitteriness, and tiredness. I am reading a book on leadership and there was one part that we shouldn't focus on being well-rounded but. 5 music documentaries that we're watching April 28. The current state of the world may have taken away being able to attend a concert in real life.

Backstage Boogie

• A multi-dimensional data model, so that data can be sliced and dicedat will, along dimensions like instance, service, endpoint, and method.
• Operational simplicity, so that you can spin up a monitoring serverwhere and when you want, even on your local workstation, withoutsetting up a distributed storage backend or reconfiguring the world.
• Scalable data collection and decentralized architecture, so that youcan reliably monitor the many instances of your services, andindependent teams can set up independent monitoring servers.
• Finally, a powerful query language that leverages the data model formeaningful alerting (including easy silencing) and graphing (fordashboards and for ad-hoc exploration).

Backstage Logic

Backstage Blogi

Backstage Login On Running

Backstage Logo

All of these features existed in various systems. However, we could notidentify a system that combined them all until a colleague started an ambitiouspet project in 2012 that aimed to do so. Shortly thereafter, we decided todevelop it into SoundCloud’s monitoring system:Prometheus was born.