signalflowcli 1.2.6

Installation

To install, along with the required dependencies, simply use pip:

$ pip install git+https://github.com/signalfx/signalflow-cli

Demo

Usage

The signalflow command-line tool supports both interactive and non-interactive operation modes. In interactive mode, it offers a prompt that allows for setting computation parameters and executing SignalFlow programs while visualizing their streaming output in real-time. The default display shows the live data, with a spark line of the last 10 values, for each received time series. CSV output and graph output can also be optained by setting the .output parameter (see below “Interactive mode usage”).

In non-interactive mode, signalflow reads the SignalFlow program text (either from stdin or from a file) and outputs the results the format specified by the --output command-line flag. By default, this uses the same live data display as the interactive prompt. The available output modes are:

• csv; outputs the data in CSV format. The first column is a millisecond timestamp; other columns contain the value of each output time series for each of those times.

  This output will keep streaming unless a fixed stopped time has been specified (which can be either in the past or in the future).

• graph; generates CSV data and renders it as a graph in a window.

• live; shows the same live display as in interactive mode, just without the prompt around it. Computation parameters should be set via the appropriate command-line flags as necessary.

Finally, the graphing can also be used from the provided standalone utility csv-to-plot, which reads CSV data from a file (or stdin) and renders the graph. Using --output graph is the same as piping the output of --output csv into csv-to-plot:

$ signalflow --start=-15m --stop=-1m --output=graph < program.txt
$ signalflow --start=-15m --stop=-1m --output=csv < program.txt | csv-to-plot

$ signalflow

-> .
{'max_delay': None,
 'output': 'live',
 'resolution': None,
 'start': '-1m',
 'stop': None}

-> .start -15m
-> .stop -1m
-> .
{'max_delay': None,
 'output': 'live',
 'resolution': None,
 'start': '-15m',
 'stop': '-1m'}

-> .stop
-> .
{'max_delay': None,
 'output': 'live',
 'resolution': None,
 'start': '-15m',
 'stop': None}

#!signalflow -x
data('cpu.utilization').mean().publish()

Obtaining your token

To obtain a session token, simply authenticate against the SignalFx API’s /v2/session endpoint as described in the API documentation: https://developers.signalfx.com/docs/session. Place the value of the sf_accessToken field in a file named .sftoken in your home directory. If you don’t have this file, or don’t use the --token parameter, the SignalFlow CLI will prompt for your username and password and obtain a token for you.