Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

fix: consume part of StreamingResponseIterator to support failure while under a retry context #10206

Merged
merged 7 commits into from
Jan 30, 2020
Merged
Show file tree
Hide file tree
Changes from 3 commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
19 changes: 19 additions & 0 deletions api_core/google/api_core/grpc_helpers.py
Original file line number Diff line number Diff line change
Expand Up @@ -65,6 +65,20 @@ class _StreamingResponseIterator(grpc.Call):
def __init__(self, wrapped):
self._wrapped = wrapped

# This iterator is used in a retry context, and returned outside after init.
# gRPC will not throw an exception until the stream is consumed, so we need
# to retrieve the first result, in order to fail, in order to trigger a retry.
try:
self._stored_first_result = six.next(self._wrapped)
except TypeError:
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@tseaver Under test we have a non-iterable. While it is odd that you would call this without it does seem possible? So added a guard here

# It is possible the wrappe method isn't an iterable (a grpc.Call
# for instance). If this happens don't store the first result.
pass
except StopIteration:

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Will this work for the WatchStream? For Watch, the client has to send the first request. We will not receive a response until after this request has been processed.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Watch uses ResumableBidiRpc and has its own way of managing recovery. This will have an effect on things like query though.

# ignore stop iteration at this time. This should be handled outside of retry.
pass


def __iter__(self):
"""This iterator is also an iterable that returns itself."""
return self
Expand All @@ -76,8 +90,13 @@ def next(self):
protobuf.Message: A single response from the stream.
"""
try:
if hasattr(self, "_stored_first_result"):
result = self._stored_first_result
del self._stored_first_result
return result
return six.next(self._wrapped)
except grpc.RpcError as exc:
# If the stream has already returned data, we cannot recover here.
six.raise_from(exceptions.from_grpc_error(exc), exc)

# Alias needed for Python 2/3 support.
Expand Down
5 changes: 2 additions & 3 deletions api_core/tests/unit/test_grpc_helpers.py
Original file line number Diff line number Diff line change
Expand Up @@ -88,6 +88,7 @@ def test_wrap_stream_iterable_iterface():

got_iterator = wrapped_callable()


callable_.assert_called_once_with()

# Check each aliased method in the grpc.Call interface
Expand Down Expand Up @@ -146,10 +147,8 @@ def test_wrap_stream_errors_iterator():

wrapped_callable = grpc_helpers._wrap_stream_errors(callable_)

got_iterator = wrapped_callable(1, 2, three="four")

with pytest.raises(exceptions.ServiceUnavailable) as exc_info:
next(got_iterator)
got_iterator = wrapped_callable(1, 2, three="four")

callable_.assert_called_once_with(1, 2, three="four")
assert exc_info.value.response == grpc_error
Expand Down