Timeouts

The client library for Python doesn't specify any default timeouts, nor are any defaults specified at the gRPC transport layer. This means that, by default, the client library for Python fully delegates timeout behavior to the server.

This is adequate for most use cases; however, if it's necessary to specify a client-side timeout, the client library for Python does support timeout overrides for both streaming and unary calls.

You can set the timeout to 2 hours or more, but the API might still time out extremely long-running requests and return a DEADLINE_EXCEEDED error. If this becomes an issue, you can split up the query and execute the chunks in parallel; this avoids the situation where a long-running request fails and the only way to recover is to restart the request.

Streaming call timeouts

The only Google Ads API service method that uses this type of call is GoogleAdsService.SearchStream.

To override the default timeout, you need to add an extra parameter when calling the method:

def make_server_streaming_call(client, customer_id):
    """Makes a server streaming call using a custom client timeout.

    Args:
        client: An initialized GoogleAds client.
        customer_id: The str Google Ads customer ID.
    """
    ga_service = client.get_service("GoogleAdsService")
    campaign_ids = []

    try:
        search_request = client.get_type("SearchGoogleAdsStreamRequest")
        search_request.customer_id = customer_id
        search_request.query = _QUERY
        stream = ga_service.search_stream(
            request=search_request,
            # When making any request, an optional "timeout" parameter can be
            # provided to specify a client-side response deadline in seconds.
            # If not set, then no timeout will be enforced by the client and
            # the channel will remain open until the response is completed or
            # severed, either manually or by the server.
            timeout=_CLIENT_TIMEOUT_SECONDS,
        )

        for batch in stream:
            for row in batch.results:
                campaign_ids.append(row.campaign.id)

        print("The server streaming call completed before the timeout.")
    except DeadlineExceeded as ex:
        print("The server streaming call did not complete before the timeout.")
        sys.exit(1)
    except GoogleAdsException as ex:
        print(
            f"Request with ID '{ex.request_id}' failed with status "
            f"'{ex.error.code().name}' and includes the following errors:"
        )
        for error in ex.failure.errors:
            print(f"\tError with message '{error.message}'.")
            if error.location:
                for field_path_element in error.location.field_path_elements:
                    print(f"\t\tOn field: {field_path_element.field_name}")
        sys.exit(1)

    print(f"Total # of campaign IDs retrieved: {len(campaign_ids)}")
set_custom_client_timeouts.py
      

Unary call timeouts

Most of the Google Ads API service methods use unary calls; typical examples are GoogleAdsService.Search and GoogleAdsService.Mutate.

To override the default timeout, you need to add an extra parameter when calling the method:

def make_unary_call(client, customer_id):
    """Makes a unary call using a custom client timeout.

    Args:
        client: An initialized GoogleAds client.
        customer_id: The Google Ads customer ID.
    """
    ga_service = client.get_service("GoogleAdsService")
    campaign_ids = []

    try:
        search_request = client.get_type("SearchGoogleAdsRequest")
        search_request.customer_id = customer_id
        search_request.query = _QUERY
        results = ga_service.search(
            request=search_request,
            # When making any request, an optional "retry" parameter can be
            # provided to specify its retry behavior. Complete information about
            # these settings can be found here:
            # https://googleapis.dev/python/google-api-core/latest/retry.html
            retry=Retry(
                # Sets the maximum accumulative timeout of the call; it
                # includes all tries.
                deadline=_CLIENT_TIMEOUT_SECONDS,
                # Sets the timeout that is used for the first try to one tenth
                # of the maximum accumulative timeout of the call.
                # Note: This overrides the default value and can lead to
                # RequestError.RPC_DEADLINE_TOO_SHORT errors when too small. We
                # recommend changing the value only if necessary.
                initial=_CLIENT_TIMEOUT_SECONDS / 10,
                # Sets the maximum timeout that can be used for any given try
                # to one fifth of the maximum accumulative timeout of the call
                # (two times greater than the timeout that is needed for the
                # first try).
                maximum=_CLIENT_TIMEOUT_SECONDS / 5,
            ),
        )

        for row in results:
            campaign_ids.append(row.campaign.id)

        print("The unary call completed before the timeout.")
    except DeadlineExceeded as ex:
        print("The unary call did not complete before the timeout.")
        sys.exit(1)
    except GoogleAdsException as ex:
        print(
            f"Request with ID '{ex.request_id}' failed with status "
            f"'{ex.error.code().name}' and includes the following errors:"
        )
        for error in ex.failure.errors:
            print(f"\tError with message '{error.message}'.")
            if error.location:
                for field_path_element in error.location.field_path_elements:
                    print(f"\t\tOn field: {field_path_element.field_name}")
        sys.exit(1)

    print(f"Total # of campaign IDs retrieved: {len(campaign_ids)}")
set_custom_client_timeouts.py