Julien Viet
Julien Viet
> Can we overcome this? Maybe not but if you can make the grpc workers be on it's own pool, that would prevent a front-end REST call from interfering with...
> Also - is quarkus really going to move away from a netty/two server option? The warning message when using a separate server is clear on this - but the...
I think we need to improve the new Vert.x gRPC stack to match the expectations from community
It seems like a good idea to increase the default window size of an HTTP/2 server, that being said we need to study the overall impact of this setting for...
I think having two separate servers could also be implemented with vertx server. in vertx we are planning to discontinue grpc-java in vertx 6 and we should actually switch in...
I think by unification beyond the technical networking aspect, we also mean that we support a unified stack where we can apply security for rest or grpc, so it makes...
@krickert I think that most configuration lies down to configured HTTP/2 server in Quarkus, unless I am wrong? what could be done in Vert.x is to have better sensitive defaults...
> [@vietj](https://github.com/vietj) - so is the plan to change the windows size only through code or will it be a configuration option? I think there should be in 5.1 an...
this should have the same implementation than `CaseInsensitiveHeaders` and only change the equality check somehow.
despite my comments, i believe we should not use inheritance for this and a simple boolean on the original class that say wether it's sensitive or not is enough.