Hi Rafael, So basically the problem here is resource vs. configuration in the gRPC side, take a look:
https://stackoverflow.com/questions/54316767/how-to-handle-resource-exhausted-error-in-api-calls The actual configuration is just a generic tune we found to offer "acceptable" performance, given some tests we did in the past. You can find this configuration inside Marvin right here: https://github.com/apache/incubator-marvin/blob/develop/engine-executor/src/main/resources/application.conf https://github.com/apache/incubator-marvin/blob/06c2138e28021ea181466c6661512757502d2c19/python-toolbox/marvin_python_toolbox/engine_base/engine_base_action.py#L185 https://github.com/apache/incubator-marvin/blob/06c2138e28021ea181466c6661512757502d2c19/python-toolbox/marvin_python_toolbox/management/engine.py#L289 Each model should have a different requirement, which leads us to a different configuration of workers, parallelism, and more. I hope I've helped you. Taka Em qui, 31 de out de 2019 às 13:38, Rafael Novello < rafa.reis.nove...@gmail.com> escreveu: > Hi guys! > > Someone knows how to deal with the following error: > > [INFO] [10/31/2019 17:29:35.737] > [lendico_score-akka.actor.default-dispatcher-3] > [GenericAPI(akka://lendico_score)] RECEIVE FAILURE!!! RESOURCE_EXHAUSTED: > Concurrent RPC limit exceeded!class io.grpc.StatusRuntimeException > > I had tested the max-workers and max-rpc-workers at engine-httpserver but I > saw no difference. > > Thanks!! > Atenciosamente, > Rafael J. R. Novello > > Skype: rafael.novello > Blog: http://rafanovello.blogspot.com.br/ >
