Databricks Community

venkad · ‎08-25-2022

Hi,

I am trying to connect to databricks workspace which has IP Access restriction enabled using databricks-sql-connector. Only my Proxy server IPs are added in the allow list.

from databricks import sql
 
connection = sql.connect(
  server_hostname ='adb-random12094383.azuredatabricks.net', 
  http_path       ='/sql/1.0/endpoints/randomid', 
  access_token='<access token here>',
)
with connection.cursor() as cur:
  cur.execute('show schemas')
  print(cur.fetchall())

This fails with the below error

---------------------------------------------------------------------------
RequestError                              Traceback (most recent call last)
Input In [2], in <cell line: 1>()
----> 1 connection = sql.connect(
      2   server_hostname ='adb-2961643968016177.17.azuredatabricks.net', 
      3   http_path       ='/sql/1.0/endpoints/17a021e95a9e6f25', 
      4   access_token='dapi3d97e76ce3de34fb4ea060cf6b068e0a',
      5   http_headers= [("https_proxy", "http://proxy-dmz.intel.com:912"), ("http_proxy", "http://proxy-dmz.intel.com:912")]
      6   )
 
File ~/project/python/venv/lib/python3.8/site-packages/databricks/sql/__init__.py:50, in connect(server_hostname, http_path, access_token, **kwargs)
     47 def connect(server_hostname, http_path, access_token, **kwargs):
     48     from .client import Connection
---> 50     return Connection(server_hostname, http_path, access_token, **kwargs)
 
File ~/project/python/venv/lib/python3.8/site-packages/databricks/sql/client.py:129, in Connection.__init__(self, server_hostname, http_path, access_token, http_headers, session_configuration, catalog, schema, **kwargs)
    120 base_headers = [("User-Agent", useragent_header)] + authorization_header
    121 self.thrift_backend = ThriftBackend(
    122     self.host,
    123     self.port,
   (...)
    126     **kwargs
    127 )
--> 129 self._session_handle = self.thrift_backend.open_session(
    130     session_configuration, catalog, schema
    131 )
    132 self.open = True
    133 logger.info("Successfully opened session " + str(self.get_session_id()))
 
File ~/project/python/venv/lib/python3.8/site-packages/databricks/sql/thrift_backend.py:444, in ThriftBackend.open_session(self, session_configuration, catalog, schema)
    435     initial_namespace = None
    437 open_session_req = ttypes.TOpenSessionReq(
    438     client_protocol_i64=ttypes.TProtocolVersion.SPARK_CLI_SERVICE_PROTOCOL_V5,
    439     client_protocol=None,
   (...)
    442     configuration=session_configuration,
    443 )
--> 444 response = self.make_request(self._client.OpenSession, open_session_req)
    445 self._check_initial_namespace(catalog, schema, response)
    446 self._check_protocol_version(response)
 
File ~/project/python/venv/lib/python3.8/site-packages/databricks/sql/thrift_backend.py:373, in ThriftBackend.make_request(self, method, request)
    371 error_info = response_or_error_info
    372 # The error handler will either sleep or throw an exception
--> 373 self._handle_request_error(error_info, attempt, elapsed)
 
File ~/project/python/venv/lib/python3.8/site-packages/databricks/sql/thrift_backend.py:241, in ThriftBackend._handle_request_error(self, error_info, attempt, elapsed)
    236     network_request_error = RequestError(
    237         user_friendly_error_message, full_error_info_context, error_info.error
    238     )
    239     logger.info(network_request_error.message_with_context())
--> 241     raise network_request_error
    243 logger.info(
    244     "Retrying request after error in {} seconds: {}".format(
    245         error_info.retry_delay, full_error_info_context
    246     )
    247 )
    248 time.sleep(error_info.retry_delay)
 
RequestError: Error during request to server

I also tried with the proxy setting along with the http_headers

  connection = sql.connect(
  server_hostname ='adb-random12094383.azuredatabricks.net', 
  http_path       ='/sql/1.0/endpoints/randomid', 
  access_token='<access token here>',
  http_headers= [("https_proxy", "proxy.mydomain.com:912"), ("http_proxy", "proxy.mydomain.com:912")]
  )

Still it fails with the same error.

How to pass the proxy setting to the Databricks Python SQL connector?

Debayan · ‎08-25-2022

Hi,

You can pass the proxy settings following the steps mentioned here: https://docs.databricks.com/dev-tools/python-sql-connector.html#insert-data

Also, would like to ask are you not importing the OS modules?

venkad · ‎08-26-2022

Sorry. I couldn't understand it. The URL shared doesn't say anything about the proxy.

Do you mean to set the proxy with the os environment variable?

I couldn't set the proxy setting on the environment level. I want to set it only for this databricks SQL module alone.

susodapop · ‎08-26-2022

`databricks-sql-connector` doesn't support HTTP proxies yet but the work is underway to implement it. Should be available in the next month or so. You can follow this issue on the open source repository for updates.