Client:
* New ManagedChannel abstract class.
* Adding ping to Channel.
* Moving builders and implementations to internal.
Server:
* Added lifecycle management API to Server (mirroring ManagedChannel).
* Moved ServerImpl, AbstractServerBuilder and handler registries to internal.
* New ServerBuilder abstract class (mirroring ManagedChannelBuilder).
Fixes#545
Using a JWT is a bit more work than it should be, but improving that
will come later.
At present, this test fails, but it is believed to be due to the auth
library.
Reserve io.grpc for public API only, and all internal stuff in core to
io.grpc.internal, including the non-stable transport API.
Raise the netty/okhttp/inprocess subpackages one level up to io.grpc,
because they are public API and entry points for most users.
Details:
- Rename io.grpc.transport to io.grpc.internal;
- Move SharedResourceHolder and SerializingExecutor to io.grpc.internal
- Rename io.grpc.transport.{netty|okhttp|inprocess} to
io.grpc.{netty|okhttp|inprocess}
This makes the reconfiguration code more concise.
- Remove configureNewStub().
- Add mutation methods withDeadlineNanoTime(), withChannel() etc that
returns the reconfigured stub.
- Remove blockingClientStreamingCall() which is not used, and we don't
actually want that API.
- Rename duplexStreamingCall() to asyncDuplexStreamingCall() to align
with other async methods.
- In unary call and client streaming call, do not request for additional
response after the first response.
Many of the interop tests were designed with the default 64 KB flow
control window in mind. If we test with 1 MB then it defeats the flow
control testing.
65 KB is an arbitrary number, but I chose it to be rather small, but
still not the HTTP/2-default 64 KB because implementations have had
trouble with applying flow control changes correctly.
This gives us more flexibility in API changes in the future.
Unary call and server streaming call should call the flow-control method
call.request() only once. Previously it was called whenever a request
arrives, which is wrong. Now it's fixed.
Resolves#436
Resolves#511.
- In generated code, make CONFIG private and METHOD_* fields public.
METHOD_* fields are MethodDescriptors now, users of the CONFIG field
should switch to using the METHOD_* fields.
- Move MethodType into MethodDescriptor (#529).
- Unify the fully qualified method name. It is fully qualified service
name + slash + short method name. It doesn't have the leading slash.
- HandlerRegistry switches the key from short method name to fully
qualified method name.
- Pass CallOptions to Channel.newCall() and
ClientInterceptor.interceptCall().
- Remove timeout from AbstractStub.StubConfigBuilder and add deadline,
which is stored in a CallOptions inside the stub.
- Deadline is in nanoseconds in the clock defined by System.nanoTime().
It is converted to timeout before transmitting on the wire. Fail the
call with DEADLINE_EXCEEDED if it's already expired.
10 years exceeds the maximum for some systems at the moment. Change to 1
year to workaround such systems while they get changed. And since "have
the default be a large number instead of not present" is a temporary fix
anyway.
Stubs don't have any timeout. However, MethodDescriptor does and
requires a timeout. We really want "no timeout," which is infinite, but
we use 10 years as the next best thing. CallOptions will end up fixing
this hack as MethodDescriptor will no longer have timeout.
Before, the hard-coded 1s didn't matter, because nobody was observing
it. Since 77878a0 it is now being sent to servers, and some servers
enforce it. Oops.
nmittler
Xudong Ma
Eric Anderson
David P. Baker
Kun Zhang
Xiao Hang
Louis Ryan
Carl Mastrangelo
Jack Coughlin