mosaik-influxdb2

This package contains an adapter to write data from a mosaik simulation into an InfluxDB 2 database.

Installation

This package is on pypi, so you can install it using pip:

pip install mosaik-influxdb2

Usage

To use the simulator, first add it to your sim_config:

sim_config = {
    "InfluxWriter": {"python": "mosaik_components.influxdb2.writer:Simulator"},
    # ...
}

Next, you need to start the simulator. Here, you have two choices to make:

1. The simulator can run both in time-based mode with a fixed step size or in event-based mode without a step size. You can choose the time-based mode by giving the parameter step_size when starting the simulator. If you give step_size=None (or don’t specify anything), the simulator will use the event-based mode.
2. You can either supply a start date (as a string parseable by Python’s datetime module) which will be combined with the (mosaik) time and time resolution to calculate each step’s time, or you can supply the time for each step on the local_time attribute (again, as a string). If you give both, the value on the local_time attribute will take precedence.

So one possible invocation would be

influx_sim = world.start("InfluxWriter",
    step_size=900,
    start_date="2022-01-01 00:00:00Z",
)

to start the simulator in time-based mode with a step size of 900 and times based on the given start date. You can leave off either argument with the effects described above.

Finally, the model needs to be started with your Influx credentials:

influx = influx_sim.Database(
    url="http://localhost:8086",
    org='.',
    bucket='my_bucket',
    token='secret_token',
    measurement='experiment_0001'
)

We recommend setting a new value for the measurement on each simulation run. (For example, you can use the start time of your simulation or a random UUID.)

Afterwards, you can define world.connect(other_entity, influx, attrs) as you like.

The simulator supports only one instance of the Database model. If you want to connect to several databases, you will need to start several instances of the simulator as well.

Other settings

By default, the adapter will save the simulator ID and entity ID of the producer of each data point in the src_sim and src_entity tags on that data point. You can change the name of these tags by passing the src_sim_tag and src_entity_tag to the start call. Setting them to None will disable writing the corresponding tag entirely.

Additionally, you can use the extra method register_src_types on the influx_sim object to pass a dict mapping entities' full IDs to arbitrary strings. Entities for which the full ID is registered in this way will have their data point augmented by the tag src_type, with the corresponding value in the dict as the tag value.

You would usually use this to store an entities "type" (for example, bus, line, PV system, etc.) as the adapter cannot reliably deduce this from the entity IDs. (Not every simulator uses the convention "ModelName-Index" for those.)

The name of this tag can be changed via the src_type_tag parameter to start.