“Cortex 1.0” Released, a Prometheus-based metrics collection system for scalable clouds
April 2, Grafana Labs released “Cortex 1.0”, a Prometheus-based metric collection system. It provides support for collecting and analyzing long-term metrics and can be used for capacity planning.
Grafana Labs is the company behind well-known projects like Grafana, and it is the leading contributor to Cortex. Cortex is a project started by Grafana Labs developers, and was accepted into the Cloud Native Computing Foundation (CNCF) Sandbox in 2018.
Having monitoring technology Prometheus as its base, Cortex is added with features like handling scalable clusters and storing obtained metrics data in the cloud storage. Cortex architecture is based on components such as Distributor (which handles automated data duplication and sending), Ingester (which is responsible for writing data sent from the Distributor to a storage backend), Ruler (which executes the rules and sends alert notifications to the Alertmanager), and Querier (which handles PromQL queries from the client). By using Prometheus Remote Write API, it handles chronicle data gathered by Prometheus instances.
It has other notable features such as speedy query handling through parallel processing and caching, Multi-Tenant, high scalability, and high availability. It supports various storage systems like Amazon Web Services (AWS) Dynamo DB, AWS S3, Apache Cassandra, and Google Cloud Bigtable. It can also store Prometheus metrics data to run capacity planning and performance analysis.
Grafana Labs has already been using Cortex for Prometheus Cloud backend. There are others who use Cortex, including Electronic Arts and Platform 9.
Cortex 1.0 is available on the project website. It is licensed under Apache License 2.