gRPC Java's transport is slightly more pluggable than C++. Is binder just inprocess? If so, why not use InProcessChannelBuilder?
On Tuesday, July 11, 2017 at 7:26:27 PM UTC-7, [email protected] wrote: > > Hi, > > Thanks for your information. > > 1. why is that possible for gRPC Java not for gRPC C++ ? > > 2. we are using the gRPC as the IPC mechanism to provide/invoke functions > between processes, binder is designed to avoid un-necessary data copy, > which have a relatively better performance then gRPC: > > e.g. writing 1k data > gRPC: 1-1.92ms > binder: 0.5ms > > If we want to make our own transport layer, where could I start from ? > Many thanks. > > Thanks, > Song > > On Wednesday, July 12, 2017 at 9:11:57 AM UTC+8, Carl Mastrangelo wrote: >> >> In gRPC Java this is possible (though poorly documented). You would need >> to make your own transport layer. >> >> What kind of performance level were you targeting? gRPC is already >> pretty fast. >> >> On Sunday, July 9, 2017 at 8:57:18 PM UTC-7, [email protected] wrote: >>> >>> Hi, >>> >>> In the Andriod platform, it provides the binder support from kernel to >>> be as an efficient transport layer compared as socket. >>> >>> So is that possible to replace the socket as binder for gRPC in Android ? >>> >>> Any suggestions for that ? many thanks in advanced. >>> >>> Thanks, >>> Song >>> >> -- You received this message because you are subscribed to the Google Groups "grpc.io" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at https://groups.google.com/group/grpc-io. To view this discussion on the web visit https://groups.google.com/d/msgid/grpc-io/64b07522-1096-4b6f-86d3-945b556b97d1%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
