[Model Monitoring] Add TDEngine Connector

[Eyal-Danieli]

Following #5210 , this PR implements TDEngine connector as a supported TSDB solution for storing model monitoring real time data.

How to define a new tdengine connection
To define a new TDEngine connection, the user can pass tsdb_connection through project.set_model_monitoring_credentials(tsdb_connection="taosws://...").

Using one of the above suggested approaches, the TDEngine connection will be stored as a k8s secret under the project and the relevant workers (stream, writer).

How the TDEngine is implemented

• The new connector is defined under mlrun/model_monitoring/db/tsdb/tdengine/tdengine_connector.py. The main connector object called TDEngineConnector and it implements the required functionality as defined in the connector class TSDBConnector.
• The implementation is based on the recommended TDEngine websocket connection. We used the TDEngine official connector taosws.
• The supertables schemas and the queries are defined under mlrun/model_monitoring/db/tsdb/tdengine/schemas.py. Each supertable should include a name, columns, tags, and a database. At the moment we have 3 different tables under the same default database - mlrun_model_monitoring. The schema design of each supertable is described below.
• The supertables are created on the 1st api call to establish the model monitoring infrastructure. Note that it only happens once because the supertables are similar accross different projects.

A related JIRA: https://iguazio.atlassian.net/browse/ML-6261

[jond01]

Requesting changes because of the confusing defaults

The changes were made