Installation

Install using pip:

pip install oml4py

or download the package and run:

python setup.py install

Usage

This module provides the OMLBase class, which contains the following methods:

• init

• start

• addmp

• inject

• close

To use OML in a python project, import the OMLBase class:

from oml4py import OMLBase

Start by initializing an OMLBase object. The init method takes up to four arguments:

• the name of the application,

• the name of the experiment,

• the name of the node,

• and the OML server URI in the form tcp:hostname:port

For example:

x=OMLBase("app", "an-exp","r","tcp:myomlserver.com:3003")

The only mandatory argument is the first one (the name of the application). If you skip any of the others, they may be defined by environment variables (OML_EXP_ID, OML_NAME, OML_SERVER). If these environment variables are not defined and they are not passed in explicitly, then the application will run with OML disabled, and the measurements that would have been sent to OML will be printed on stdout instead.

Next, add one or more measurement points. Pass the name of the measurement point and a schema string to the start method. The schema string should be in the format measurement_name1:measurement_type1 measurement_name2:measurement_type2 for example:

x.addmp("fft", "freq:long amplitude:double fft_val:double")

When you have set up all your measurement points, call start():

x.start()

When you have a measurement point to send to OML, store them in a list. Then pass the name of the measurement point and list of values to inject, as follows:

x.inject("fft", [259888, 15, -38])

At the end of your program, call close to gracefully close the database:

x.close()