AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |
Back to Blog
Prometheus cloudwatch exporter example12/31/2023 ![]() StatsD is a network daemon built in Node.js that listens for messages containing statistics, like counters and timers, so that it can send aggregates to configurable backends.įigure 1: DoorDash previously used StatsD proxy and server pipelines for microservices’ metrics StatsD was originally developed and released by Etsy to aggregate and summarize application metrics. We’ll briefly introduce StatD’s history before diving into those specific issues. StatsD was a great asset for our early observability needs, but we began encountering constraints such as losing metrics during surge events, difficulties with naming/standardized tags, and a lack of reporting tools. In this article, we explain how we accomplished the migration, first reviewing the challenges we faced with StatsD and then discussing how we selected and implemented Prometheus as our preferred solution. As we continue to scale DoorDash’s business, using Prometheus lets us eliminate metrics loss, scale up our metrics usage, standardize our metrics labels, and significantly lower overall costs. That’s why we decided to migrate our observability technology stack to Prometheus-based monitoring. Just when we most needed observability data, the system would leave us in the lurch. ![]() Unfortunately, this was a challenge at DoorDash because of peak traffic failures while using our legacy metrics infrastructure based on StatsD. Accurate and reliable observability is essential when supporting a large distributed service, but this is only possible if your tools are equally scalable.
0 Comments
Read More
Leave a Reply. |