Hastic server

Website | Twitter

Implementation of basic pattern recognition for anomaly detection.

Implementation of analytics unit for Hastic.

Please note that we are still in alpha, so features are subject to change

See also:

• Hooks - notifications about events
• REST - for developing your plugins
• HasticPanel - Hastic visualisation plugin for Grafana

Download & Install on Linux

You need only nodejs >= 6.14 on your machine.

Node 6

wget https://github.com/hastic/hastic-server/releases/download/0.2.2-alpha/hastic-server-0.2.2-alpha-node-6.tar.gz
tar -zxvf hastic-server-0.2.2-alpha-node-6.tar.gz
cd hastic-server-0.2.2-alpha/server/dist
node server

Node 8

wget https://github.com/hastic/hastic-server/releases/download/0.2.2-alpha/hastic-server-0.2.2-alpha-node-8.tar.gz
tar -zxvf hastic-server-0.2.2-alpha-node-8.tar.gz
cd hastic-server-0.2.2-alpha/server/dist
node server

Build & run from source

Hastic server requires Grafana's API key (http://<your_grafana_url>/org/apikeys) to query data from Grafana datasources. API key role requires only Viewer access.

Possible to install on:

• Linux
• Docker

Linux

System prerequisites:

• git
• nodejs >= 6.14
• python == 3.6.х with pip3

Installation

git clone https://github.com/hastic/hastic-server.git
cd hastic-server
pip3 install -r analytics/requirements.txt
cd server
npm install
npm run build

Configuration

You can configure hastic-server using either environment variables or config file.

NOTE: environment variables have higher priority than config file.

Environment variables

You can export the following environment variables for hastic-server to use:

• HASTIC_API_KEY - (required) API-key of your Grafana instance
• HASTIC_PORT - (optional) port you want to run server on, default: 8000

e.g.

export HASTIC_API_KEY=eyJrIjoiVjZqMHY0dHk4UEE3eEN4MzgzRnd2aURlMWlIdXdHNW4iLCJuIjoiaGFzdGljIiwiaWQiOjF9
export HASTIC_PORT=8080

Config file

You can also rename config.example.json to config.json and set your values there.

Run

cd hastic-server/server
npm start

Docker

Fetch image

docker pull hastic/server

Run

docker run -d \
  --name hastic-server \
  --ipc host \
  -p 80:8000 \
  -e HASTIC_API_KEY=<your_grafana_api_key> \
  -v /tmp:/tmp \
  hastic/server

Make sure, that HASTIC_PORT is opened in your system. Default value is 8000.

Changelog

[0.2.2-alpha] - 2018-09-12

NOTE: hastic-panels of versions older than 0.2.0 are not supported

Changed

• Consider segment width in models #136
• Increase learning dataset timerange

[0.2.1-alpha] - 2018-09-04

NOTE: hastic-panels of versions older than 0.2.0 are not supported

Changed

• Models improvements

[0.2.0-alpha] - 2018-09-03

NOTE: hastic-panels of versions older than 0.2.0 are not supported

Added

• Return version of server #66
• Return analytic status #71

Fixed

• Case-sensitive anomaly name #41
• ImportError: cannot import name 'isna' #59
• Handle analytics connection failure on server #74
• Analytics wont start in production mode #77
• Basic creation of analytic unit fails #79
• Missing communication between analytics and server in production #91 - thx @petrk94 for bugreport
• Server is stopping only after second Ctrl-C #96
• Error: 'value' #100
• Error: Unexpected response #125
• Error: max of an empty sequence in peaks model #126

[0.1.4-alpha] - 2018-06-29

Changed

• Informative error messages instead of "Internal error" #40

Fixed

• "No such file or directory" error on anomaly create #33
• Case-sensitive anomaly name #41

[0.1.3-alpha] - 2018-06-28

Changed

• Drops algorithm improvement.

[0.1.2-alpha] - 2018-06-25

Fixed

• Error: type object 'sklearn.tree...' #28

[0.1.1-alpha] - 2018-06-25

Added

• HASTIC_API_KEY to config file #23