Beringei is an embedded in-memory storage engine for time series data developed by Facebook. It was developed by Facebook to handle the growing scale of their system health monitoring data, as they needed a faster solution to storing constant stream of time series data, as well as a fast way to query this data.[01]
- Website
- https://engineering.fb.com/2017/02/03/core-infra/beringei-a-high-performance-time-series-storage-engine/[01]
- Source Code
- https://github.com/facebookarchive/beringei[02]
- Developer
- Country of Origin
- US
- Start Year
- 2015
- End Year
- 2017
- Former Name
- Gorilla
- Project Type
- Open Source
- Written in
- C++
- Supported Languages
- C++
- Derived From
- GOLD STAR
- Operating System
- Linux
- License
- BSD License
Beringei is an embedded in-memory storage engine for time series data developed by Facebook. It was developed by Facebook to handle the growing scale of their system health monitoring data, as they needed a faster solution to storing constant stream of time series data, as well as a fast way to query this data.[01]
History[03]
The original name of the system was called Gorilla. The architecture was described in a VLDB 2015 research paper. The project was then open-sourced in 2017 and then discontinued in 2018. The original solution to handling monitoring data was an HBase-backed TSDB (Time series database), but the primary issues revolved around slow aggregate queries, and slow writes.
