Benchmarks for the data sourcing actor

[shsms]

What's needed?

We need benchmarks for the data sourcing actor, where we see how long
it takes to stream a certain number of values for a certain number of
microgrid components.

Proposed solution

Prereqs:

1. The MockMicrogrid package in tests currently sends out 2000
   values per component_id. This needs to be configurable.

2. The start method starts the data sourcing actor and the
   resampling actor, but sends out the channel only for the resampling
   actor. There should ideally be a start_mock_client method that
   just sets up the mock client, which the start method calls. Then
   we would be able just call start_mock_client and then create our
   own actor instances.

The data sourcing actor can be benchmarked by a function that takes as
input:

• number of components
• the number of messages to be sent out for each battery as inputs

This benchmarking function can be implemented like this:

1. Create a MockMicrogrid instance, with the given number of ev
   chargers like here.
2. the MockMicrogrid has a default sample_rate_s of 0.01, this can
   be set to 0.0 (or almost 0.0 if there are synchronization issues).
3. Create a DataSourcingActor instance, and send requests to it for
   each of the 3 current phases,
   for all component ids in MockMicrogrid.evc_ids. Collect the
   corresponding receivers into a channels.MergeNamed and count the
   number of messages received from each component_id+metrid_id pair.
4. Confirm the number of messages sent is the number of messages
   received. Measure the time required to stream the configured
   number of messages.

Use cases

No response

Alternatives and workarounds

No response

Additional context

No response

[mathias-baumann-frequenz]

Should step 1. also add batteries and not just ev chargers?