cw2018 / banshee

Anomalies detection system for periodic metrics.

Banshee

Banshee is a real-time anomalies(outliers) detection system for periodic metrics.

Requirements

1. Go >= 1.4 and godep.
2. Node and gulp.

Build

1. Clone the repo.
2. Build binary via make.
3. Build static files via make static.

Philosophy

3-sigma:

>>> import numpy as np
>>> x = np.array([40, 52, 63, 44, 54, 43, 67, 54, 49, 45, 48, 54, 57, 43, 58])
>>> mean = np.mean(x)
>>> std = np.std(x)
>>> (80 - mean) / (3 * std)
1.2608052883472445 # anomaly, too big
>>> (20 - mean) / (3 * std)
-1.3842407711224991 # anomaly, too small

Documentation

https://godoc.org/github.com/eleme/banshee

Statsd Backend

https://www.npmjs.com/package/statsd-banshee

Migrate from bell

https://godoc.org/github.com/eleme/banshee#hdr-Migrate_from_bell

Authors

• @facesea
• @nrzipher
• @hit9

License

MIT Copyright (c) 2015 - 2016 Eleme, Inc.