Hyperpilotio/analyzer

Analyzer that performs configuration searching and metrics correlation

Analyzer

Analyzer performs interference prediction and validation and visualizes the data.

Usage:

1. Build docker image

    make docker-build
   

2. Run server

    make docker-run
   

3. Test server

    make docker-test
   

Configuration:

Configuration file is located in 'config.ini'

Prerequisite:

In order to connect to database, connecting to VPN is needed for now.

UI:

make docker-build
make docker-run

then connect to: http://localhost:5000/api

Setting up and running the app (without Docker)

• make init
• make serve (don't forget to connect to VPN first)

Please install Python 3.6.1 or later if you see the message from running make init saying that the Python version is outdated.

Diagnosis

Diagnosis is our end-to-end node and container problem detection product.

Configuration:

Configuration file is located in 'config.ini'

Requirements:

1. Analyzer API and pipenv

• Follow instruction for setting up pipenv. (https://github.com/Hyperpilotio/analyzer/blob/master/api_service/README.md)

• After running pipenv shell

  run the API from the analyzer directory with make dev

2. Influx

• Install influx and run the server using the command influxd.

• Install and configure aws. After installation, run aws configure and use the shared amazon s3 credentials here: (https://github.com/Hyperpilotio/hyperpilot-demo/wiki)

• From the hyperpilot-demo repo, run /hyperpilot_influx_restore.sh -n {name_of_backup_file} to restore a snapshot uploaded to s3 to your local influx.

3. Mongo

• Install and run server with mongod.

• From the mongo_service directory, run mongo create_user.js.

Running the analysis:

• From the analyzer directory (with the API, influxd, mongod and activated pipenv) run python -u -m diagnosis.app_analyzer
• Derived metrics and diagnosis results will be written to new influx databases. Problems and other diagnosis-related collections will be written to mongo.