grpc/testing/control.proto

// Copyright 2015 gRPC authors.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
//     http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.

syntax = "proto3";

import "src/proto/grpc/testing/payloads.proto";
import "src/proto/grpc/testing/stats.proto";
import "google/protobuf/timestamp.proto";

package grpc.testing;

enum ClientType {
  // Many languages support a basic distinction between using
  // sync or async client, and this allows the specification
  SYNC_CLIENT = 0;
  ASYNC_CLIENT = 1;
  OTHER_CLIENT = 2; // used for some language-specific variants
  CALLBACK_CLIENT = 3;
}

enum ServerType {
  SYNC_SERVER = 0;
  ASYNC_SERVER = 1;
  ASYNC_GENERIC_SERVER = 2;
  OTHER_SERVER = 3; // used for some language-specific variants
  CALLBACK_SERVER = 4;
}

enum RpcType {
  UNARY = 0;
  STREAMING = 1;
  STREAMING_FROM_CLIENT = 2;
  STREAMING_FROM_SERVER = 3;
  STREAMING_BOTH_WAYS = 4;
}

// Parameters of poisson process distribution, which is a good representation
// of activity coming in from independent identical stationary sources.
message PoissonParams {
  // The rate of arrivals (a.k.a. lambda parameter of the exp distribution).
  double offered_load = 1;
}

// Once an RPC finishes, immediately start a new one.
// No configuration parameters needed.
message ClosedLoopParams {}

message LoadParams {
  oneof load {
    ClosedLoopParams closed_loop = 1;
    PoissonParams poisson = 2;
  };
}

// presence of SecurityParams implies use of TLS
message SecurityParams {
  bool use_test_ca = 1;
  string server_host_override = 2;
  string cred_type = 3;
}

message ChannelArg {
  string name = 1;
  oneof value {
    string str_value = 2;
    int32 int_value = 3;
  }
}

message ClientConfig {
  // List of targets to connect to. At least one target needs to be specified.
  repeated string server_targets = 1;
  ClientType client_type = 2;
  SecurityParams security_params = 3;
  // How many concurrent RPCs to start for each channel.
  // For synchronous client, use a separate thread for each outstanding RPC.
  int32 outstanding_rpcs_per_channel = 4;
  // Number of independent client channels to create.
  // i-th channel will connect to server_target[i % server_targets.size()]
  int32 client_channels = 5;
  // Only for async client. Number of threads to use to start/manage RPCs.
  int32 async_client_threads = 7;
  RpcType rpc_type = 8;
  // The requested load for the entire client (aggregated over all the threads).
  LoadParams load_params = 10;
  PayloadConfig payload_config = 11;
  HistogramParams histogram_params = 12;

  // Specify the cores we should run the client on, if desired
  repeated int32 core_list = 13;
  int32 core_limit = 14;

  // If we use an OTHER_CLIENT client_type, this string gives more detail
  string other_client_api = 15;

  repeated ChannelArg channel_args = 16;

  // Number of threads that share each completion queue
  int32 threads_per_cq = 17;

  // Number of messages on a stream before it gets finished/restarted
  int32 messages_per_stream = 18;

  // Use coalescing API when possible.
  bool use_coalesce_api = 19;

  // If 0, disabled. Else, specifies the period between gathering latency
  // medians in milliseconds.
  int32 median_latency_collection_interval_millis = 20;

  // Number of client processes. 0 indicates no restriction.
  int32 client_processes = 21;
}

message ClientStatus { ClientStats stats = 1; }

// Request current stats
message Mark {
  // if true, the stats will be reset after taking their snapshot.
  bool reset = 1;
}

message ClientArgs {
  oneof argtype {
    ClientConfig setup = 1;
    Mark mark = 2;
  }
}

message ServerConfig {
  ServerType server_type = 1;
  SecurityParams security_params = 2;
  // Port on which to listen. Zero means pick unused port.
  int32 port = 4;
  // Only for async server. Number of threads used to serve the requests.
  int32 async_server_threads = 7;
  // Specify the number of cores to limit server to, if desired
  int32 core_limit = 8;
  // payload config, used in generic server.
  // Note this must NOT be used in proto (non-generic) servers. For proto servers,
  // 'response sizes' must be configured from the 'response_size' field of the
  // 'SimpleRequest' objects in RPC requests.
  PayloadConfig payload_config = 9;

  // Specify the cores we should run the server on, if desired
  repeated int32 core_list = 10;

  // If we use an OTHER_SERVER client_type, this string gives more detail
  string other_server_api = 11;

  // Number of threads that share each completion queue
  int32 threads_per_cq = 12;

  // c++-only options (for now) --------------------------------

  // Buffer pool size (no buffer pool specified if unset)
  int32 resource_quota_size = 1001;
  repeated ChannelArg channel_args = 1002;

  // Number of server processes. 0 indicates no restriction.
  int32 server_processes = 21;
}

message ServerArgs {
  oneof argtype {
    ServerConfig setup = 1;
    Mark mark = 2;
  }
}

message ServerStatus {
  ServerStats stats = 1;
  // the port bound by the server
  int32 port = 2;
  // Number of cores available to the server
  int32 cores = 3;
}

message CoreRequest {
}

message CoreResponse {
  // Number of cores available on the server
  int32 cores = 1;
}

message Void {
}

// A single performance scenario: input to qps_json_driver
message Scenario {
  // Human readable name for this scenario
  string name = 1;
  // Client configuration
  ClientConfig client_config = 2;
  // Number of clients to start for the test
  int32 num_clients = 3;
  // Server configuration
  ServerConfig server_config = 4;
  // Number of servers to start for the test
  int32 num_servers = 5;
  // Warmup period, in seconds
  int32 warmup_seconds = 6;
  // Benchmark time, in seconds
  int32 benchmark_seconds = 7;
  // Number of workers to spawn locally (usually zero)
  int32 spawn_local_worker_count = 8;
}

// A set of scenarios to be run with qps_json_driver
message Scenarios {
  repeated Scenario scenarios = 1;
}

// Basic summary that can be computed from ClientStats and ServerStats
// once the scenario has finished.
message ScenarioResultSummary
{
  // Total number of operations per second over all clients. What is counted as 1 'operation' depends on the benchmark scenarios:
  // For unary benchmarks, an operation is processing of a single unary RPC.
  // For streaming benchmarks, an operation is processing of a single ping pong of request and response.
  double qps = 1;
  // QPS per server core.
  double qps_per_server_core = 2;
  // The total server cpu load based on system time across all server processes, expressed as percentage of a single cpu core.
  // For example, 85 implies 85% of a cpu core, 125 implies 125% of a cpu core. Since we are accumulating the cpu load across all the server
  // processes, the value could > 100 when there are multiple servers or a single server using multiple threads and cores.
  // Same explanation for the total client cpu load below.
  double server_system_time = 3;
  // The total server cpu load based on user time across all server processes, expressed as percentage of a single cpu core. (85 => 85%, 125 => 125%)
  double server_user_time = 4;
  // The total client cpu load based on system time across all client processes, expressed as percentage of a single cpu core. (85 => 85%, 125 => 125%)
  double client_system_time = 5;
  // The total client cpu load based on user time across all client processes, expressed as percentage of a single cpu core. (85 => 85%, 125 => 125%)
  double client_user_time = 6;

  // X% latency percentiles (in nanoseconds)
  double latency_50 = 7;
  double latency_90 = 8;
  double latency_95 = 9;
  double latency_99 = 10;
  double latency_999 = 11;

  // server cpu usage percentage
  double server_cpu_usage = 12;

  // Number of requests that succeeded/failed
  double successful_requests_per_second = 13;
  double failed_requests_per_second = 14;

  // Number of polls called inside completion queue per request
  double client_polls_per_request = 15;
  double server_polls_per_request = 16;

  // Queries per CPU-sec over all servers or clients
  double server_queries_per_cpu_sec = 17;
  double client_queries_per_cpu_sec = 18;


  // Start and end time for the test scenario
  google.protobuf.Timestamp start_time = 19;
  google.protobuf.Timestamp end_time =20;
}

// Results of a single benchmark scenario.
message ScenarioResult {
  // Inputs used to run the scenario.
  Scenario scenario = 1;
  // Histograms from all clients merged into one histogram.
  HistogramData latencies = 2;
  // Client stats for each client
  repeated ClientStats client_stats = 3;
  // Server stats for each server
  repeated ServerStats server_stats = 4;
  // Number of cores available to each server
  repeated int32 server_cores = 5;
  // An after-the-fact computed summary
  ScenarioResultSummary summary = 6;
  // Information on success or failure of each worker
  repeated bool client_success = 7;
  repeated bool server_success = 8;
  // Number of failed requests (one row per status code seen)
  repeated RequestResultCount request_results = 9;
}