CP4D Audit: Ability to trace activities related to assets’ execution for the Watson Studio service (and the CP4D Platform)

CP4D Audit: Ability to trace activities related to assets’ execution for the Watson Studio service (and the CP4D Platform)

Client would like to be able to trace the activities related to the execution of an asset. With reference to the Watson Studio Project assets for instance, currently (CP4D 5.1), the auditable events include only the following:

• data-science-experience.project-asset.create – Create an asset

• data-science-experience.project-asset.patch – Rename an asset

• data-science-experience.project-asset.delete – Delete an asset

• data-science-experience.project-download.create – Download an asset

• data-science-experience.project-attachment.patch – Rename an attachment

So, there is no tracing of the actual activity performed by the user within the asset. It would be useful to trace the access to the asset, what the user did with the asset itself, if there was an external connection, what happened during such an interaction, etc.

For instance during the execution of the following code within a Notebook there is no trace of the interactions with the Flight service to read from a data source:

import itc_utils.flight_service as itcfs
# NOTE:
#  A limit of 5000 rows has been applied to the request to enable sample previewing.
#  Adjust the display message as needed by editing the following lines:
from IPython.display import display, HTML
display(HTML("A
 row limit of 5000 has been applied to the query to enable sample 
previewing. If the data set is larger, only the first 5000 rows will be 
loaded."))
#  Edit select_statement to change or disable the row limit.
#
nb_data_request = {
    'connection_name': """test bigsql""",
    'interaction_properties': {
        'select_statement': ‘select * from table’
    }
}
flight_descriptor = itcfs.get_flight_descriptor(nb_data_request=nb_data_request)
 
flightClient = itcfs.get_flight_client()
flightInfo = flightClient.get_flight_info(flight_descriptor)
 
df_0 = itcfs.read_pandas_and_concat(flightClient, flightInfo, timeout=240)
df_0.head(10)