Databricks SQL Connector

[BorislavBlagoev]

I started to explore Databricks SQL Connector. And I'm wondering can I execute query from Databricks SQL Connector from notebook or job?

sample code:

'test_query' lives in Databricks SQL. It is ready to use query!

from databricks import sql
 
hostname = '<name>.databricks.com'
http_path = '/sql/1.0/endpoints/<endpoint_id>'
access_token = '<personal_token>'
 
connection = sql.connect(server_hostname=hostname, http_path=http_path, access_token=access_token)
 
cursor = connection.cursor()
 
cursor.execute('test_query')
 
result = cursor.fetchall()
 
for row in result:
  print(row)
 
cursor.close()

[Hubert-Dudek]

yes you can as it is just python but in notebook you can use native SQL commands (of course there can be situation that you have some really big sql enpoint (like 64 DBU) so it can make sense in some situations but I don't think much and it creates complicated architecture)

[BorislavBlagoev]

Than that make more sense to use that in job? And what is the big benefit of Databricks SQL at all? I mean what is the main usage of Databricks SQL?

[Hubert-Dudek · Accepted solution]

except nice UI and cool features there 😉 for me biggest advantage is that from PowerBI and Qlikview I can query data using SQL and use serverless SQL endpoint.

[BorislavBlagoev]

That means there is no advantages if I use Databricks SQL in my job

[Hubert-Dudek]

I would use just native SQL syntax in job but maybe there is some nice trick...

[BorislavBlagoev]

Thank you!

[Anonymous]

@Borislav Blagoev​ - Would you be happy to mark which of @Hubert Dudek​'s answers offered the best resolution as best?

[BorislavBlagoev]

I can't understand your question

[BorislavBlagoev]

@Hubert Dudek​ 

Error during OpenSession; Request: TOpenSessionReq(client_protocol=5, username=None, password=None, configuration=None) Error: None Bounded-Retry-Delay: None Attempt: 1/30 Elapsed-Seconds: 1.3954582214355469/900.0
OperationalError: ('', EOFError())

I get the following error when I try to execute the code above. Do you know something about this?

[Anonymous]

@Borislav Blagoev​ - There is a "Select as Best" link underneath each reply. All you need to do, if you choose, is to click on that link for whichever reply best answered your question. The purpose is to help other members find helpful solutions more easily.

I'm attaching a screenshot to help.

[BorislavBlagoev]

Done!