Question : I am using a

dynamic partition

for a CDC load (an dbt incremental materialization). How can i “initialize” the partition have 1 partition named

INITIAL_LOAD

and then use a sensor to add further paritions ? Also, how can I record a data against this partition as my start date ?

fivetran_prod_test_connection_test_03 = DynamicPartitionsDefinition(
    # <https://docs.dagster.io/_apidocs/partitions#dagster.DynamicPartitionsDefinition>
    name="fivetran_prod_test_connection_test_03",
)
fivetran_prod_test_connection_test_03.instance.add_dynamic_partitions(fivetran_prod_test_connection_test_03.name, ['INITIAL_LOAD']) 
# Where can I add a start_date config to this partition ?

Another question: Can we specify the name of dbt ops that it created using the dbt asset? Like in this example, the dbt run name "run_dbt_3a4be". Can we config this name?

How can i “initialize” the partition have 1 partition named

INITIAL_LOAD

and then use a sensor to add further paritions ?

In a script, you could do this following:

instance = DagsterInstance.get()
instance.add_dynamic_partitions("partitions_def_name", ["INITIAL_LOAD"])

Also, how can I record a data against this partition as my start date ?

Are you able to explain what you mean by this in more detail? There isn't currently the capability to attach metadata to a partition.

Can we specify the name of dbt ops that it created using the dbt asset? Like in this example, the dbt run name "run_dbt_3a4be". Can we config this name? (edited)

This isn't currently possible. If you file a Github issue for this, we might be able to add support for it.