Fixing code review comments:
- Fixing for comments in all except for test file
- Multi-purpose test_service_impl.h as oppose
to creating new files.
diff --git a/include/grpc/impl/codegen/grpc_types.h b/include/grpc/impl/codegen/grpc_types.h
index 1e4c9fd..5a73918 100644
--- a/include/grpc/impl/codegen/grpc_types.h
+++ b/include/grpc/impl/codegen/grpc_types.h
@@ -358,7 +358,10 @@
  * The default is 15 seconds. */
 #define GRPC_ARG_XDS_RESOURCE_DOES_NOT_EXIST_TIMEOUT_MS \
   "grpc.xds_resource_does_not_exist_timeout_ms"
-/* if set, enable xds routing policy */
+/* If set, enable xds routing policy.  This boolean argument is currently
+ * disabled by default; however, it will be changed to enabled by default
+ * once the functionality proves stable.  This arg will eventually
+ * be removed completely. */
 #define GRPC_ARG_XDS_ROUTING_ENABLED "grpc.xds_routing_enabled"
 /** If non-zero, grpc server's cronet compression workaround will be enabled */
 #define GRPC_ARG_WORKAROUND_CRONET_COMPRESSION \
diff --git a/src/core/ext/filters/client_channel/lb_policy/xds/xds_routing.cc b/src/core/ext/filters/client_channel/lb_policy/xds/xds_routing.cc
index 7bcc05c..4017467 100644
--- a/src/core/ext/filters/client_channel/lb_policy/xds/xds_routing.cc
+++ b/src/core/ext/filters/client_channel/lb_policy/xds/xds_routing.cc
@@ -610,7 +610,7 @@
     std::vector<grpc_error*> error_list;
     // action map.
     XdsRoutingLbConfig::ActionMap action_map;
-    std::set<std::string /*action_name*/> action_to_be_used;
+    std::set<std::string /*action_name*/> actions_to_be_used;
     auto it = json.object_value().find("actions");
     if (it == json.object_value().end()) {
       error_list.push_back(GRPC_ERROR_CREATE_FROM_STATIC_STRING(
@@ -620,6 +620,11 @@
           "field:actions error:type should be object"));
     } else {
       for (const auto& p : it->second.object_value()) {
+        if (p.first.empty()) {
+          error_list.push_back(GRPC_ERROR_CREATE_FROM_STATIC_STRING(
+              "field:actions element error: name cannot be empty"));
+          continue;
+        }
         RefCountedPtr<LoadBalancingPolicy::Config> child_config;
         std::vector<grpc_error*> child_errors =
             ParseChildConfig(p.second, &child_config);
@@ -634,7 +639,7 @@
           error_list.push_back(error);
         } else {
           action_map[p.first] = std::move(child_config);
-          action_to_be_used.insert(p.first);
+          actions_to_be_used.insert(p.first);
         }
       }
     }
@@ -655,7 +660,7 @@
       for (size_t i = 0; i < array.size(); ++i) {
         XdsRoutingLbConfig::Route route;
         std::vector<grpc_error*> route_errors =
-            ParseRoute(array[i], action_map, &route, &action_to_be_used);
+            ParseRoute(array[i], action_map, &route, &actions_to_be_used);
         if (!route_errors.empty()) {
           // Can't use GRPC_ERROR_CREATE_FROM_VECTOR() here, because the error
           // string is not static in this case.
@@ -680,7 +685,7 @@
           "default route must not contain service or method");
       error_list.push_back(error);
     }
-    if (!action_to_be_used.empty()) {
+    if (!actions_to_be_used.empty()) {
       error_list.push_back(GRPC_ERROR_CREATE_FROM_STATIC_STRING(
           "some actions were not referenced by any route"));
     }
@@ -760,7 +765,7 @@
   static std::vector<grpc_error*> ParseRoute(
       const Json& json, const XdsRoutingLbConfig::ActionMap& action_map,
       XdsRoutingLbConfig::Route* route,
-      std::set<std::string /*action_name*/>* action_to_be_used) {
+      std::set<std::string /*action_name*/>* actions_to_be_used) {
     std::vector<grpc_error*> error_list;
     if (json.type() != Json::Type::OBJECT) {
       error_list.push_back(GRPC_ERROR_CREATE_FROM_STATIC_STRING(
@@ -790,15 +795,19 @@
           "field:action error:should be of type string"));
     } else {
       route->action = it->second.string_value();
-      // Validate action exists and mark it as used.
-      if (!route->action.empty() &&
-          action_map.find(route->action) == action_map.end()) {
+      if (route->action.empty()) {
         error_list.push_back(GRPC_ERROR_CREATE_FROM_STATIC_STRING(
-            absl::StrCat("field:action error:", route->action,
-                         " does not exist")
-                .c_str()));
+            "field:action error:cannot be empty"));
+      } else {
+        // Validate action exists and mark it as used.
+        if (action_map.find(route->action) == action_map.end()) {
+          error_list.push_back(GRPC_ERROR_CREATE_FROM_STATIC_STRING(
+              absl::StrCat("field:action error:", route->action,
+                           " does not exist")
+                  .c_str()));
+        }
+        actions_to_be_used->erase(route->action);
       }
-      action_to_be_used->erase(route->action);
     }
     return error_list;
   }
diff --git a/src/core/ext/filters/client_channel/xds/xds_api.cc b/src/core/ext/filters/client_channel/xds/xds_api.cc
index d74b2ab..855d34a 100644
--- a/src/core/ext/filters/client_channel/xds/xds_api.cc
+++ b/src/core/ext/filters/client_channel/xds/xds_api.cc
@@ -1013,11 +1013,9 @@
     return GRPC_ERROR_CREATE_FROM_STATIC_STRING(
         "No route found in the virtual host.");
   }
-
   // If xds_routing is not configured, only look at the last one in the route
   // list (the default route)
-  size_t start_index = size - 1;
-  if (xds_routing_enabled) start_index = 0;
+  size_t start_index = xds_routing_enabled ? 0 : size - 1;
   for (size_t i = start_index; i < size; ++i) {
     const envoy_api_v2_route_Route* route = routes[i];
     const envoy_api_v2_route_RouteMatch* match =
@@ -1064,6 +1062,10 @@
     }
     const upb_strview action =
         envoy_api_v2_route_RouteAction_cluster(route_action);
+    if (action.size == 0) {
+      return GRPC_ERROR_CREATE_FROM_STATIC_STRING(
+          "RouteAction has empty cluster.");
+    }
     rds_route.cluster_name = std::string(action.data, action.size);
     rds_update->routes.emplace_back(std::move(rds_route));
   }
diff --git a/src/core/ext/filters/client_channel/xds/xds_client.cc b/src/core/ext/filters/client_channel/xds/xds_client.cc
index 33e9d17..2f06574 100644
--- a/src/core/ext/filters/client_channel/xds/xds_client.cc
+++ b/src/core/ext/filters/client_channel/xds/xds_client.cc
@@ -903,7 +903,7 @@
                  ? lds_update->route_config_name.c_str()
                  : "<inlined>"));
     if (lds_update->rds_update.has_value()) {
-      gpr_log(GPR_INFO, "  RouteConfiguration contains %lu routes", this,
+      gpr_log(GPR_INFO, "  RouteConfiguration contains %lu routes",
               lds_update->rds_update.value().routes.size());
       for (const auto& route : lds_update->rds_update.value().routes) {
         gpr_log(GPR_INFO,
@@ -1819,8 +1819,8 @@
 }
 
 bool GetXdsRoutingEnabled(const grpc_channel_args& args) {
-  return grpc_channel_args_find_integer(&args, GRPC_ARG_XDS_ROUTING_ENABLED,
-                                        {0, 0, 1});
+  return grpc_channel_args_find_bool(&args, GRPC_ARG_XDS_ROUTING_ENABLED,
+                                     false);
 }
 
 }  // namespace
diff --git a/test/cpp/end2end/BUILD b/test/cpp/end2end/BUILD
index b8dbf5b..bb6be0c 100644
--- a/test/cpp/end2end/BUILD
+++ b/test/cpp/end2end/BUILD
@@ -38,7 +38,7 @@
 grpc_cc_library(
     name = "test_multiple_service_impl",
     testonly = True,
-    hdrs = ["test_multiple_service_impl.h"],
+    hdrs = ["test_service_impl.h"],
     external_deps = [
         "gtest",
     ],
diff --git a/test/cpp/end2end/test_multiple_service_impl.h b/test/cpp/end2end/test_multiple_service_impl.h
deleted file mode 100644
index 2b06117..0000000
--- a/test/cpp/end2end/test_multiple_service_impl.h
+++ /dev/null
@@ -1,495 +0,0 @@
-/*
- *
- * Copyright 2016 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.
- *
- */
-
-#ifndef GRPC_TEST_CPP_END2END_TEST_MULTIPLE_SERVICE_IMPL_H
-#define GRPC_TEST_CPP_END2END_TEST_MULTIPLE_SERVICE_IMPL_H
-
-#include <condition_variable>
-#include <memory>
-#include <mutex>
-
-#include <grpc/grpc.h>
-#include <grpc/support/log.h>
-#include <grpcpp/alarm.h>
-#include <grpcpp/security/credentials.h>
-#include <grpcpp/server_context.h>
-#include <gtest/gtest.h>
-
-#include <string>
-#include <thread>
-
-#include "src/proto/grpc/testing/echo.grpc.pb.h"
-#include "test/cpp/util/string_ref_helper.h"
-
-using std::chrono::system_clock;
-
-namespace grpc {
-namespace testing {
-
-const int kServerDefaultResponseStreamsToSend = 3;
-const char* const kServerResponseStreamsToSend = "server_responses_to_send";
-const char* const kServerTryCancelRequest = "server_try_cancel";
-const char* const kDebugInfoTrailerKey = "debug-info-bin";
-const char* const kServerFinishAfterNReads = "server_finish_after_n_reads";
-const char* const kServerUseCoalescingApi = "server_use_coalescing_api";
-const char* const kCheckClientInitialMetadataKey = "custom_client_metadata";
-const char* const kCheckClientInitialMetadataVal = "Value for client metadata";
-
-typedef enum {
-  DO_NOT_CANCEL = 0,
-  CANCEL_BEFORE_PROCESSING,
-  CANCEL_DURING_PROCESSING,
-  CANCEL_AFTER_PROCESSING
-} ServerTryCancelRequestPhase;
-
-namespace {
-
-// When echo_deadline is requested, deadline seen in the ServerContext is set in
-// the response in seconds.
-void MaybeEchoDeadline(experimental::ServerContextBase* context,
-                       const EchoRequest* request, EchoResponse* response) {
-  if (request->has_param() && request->param().echo_deadline()) {
-    gpr_timespec deadline = gpr_inf_future(GPR_CLOCK_REALTIME);
-    if (context->deadline() != system_clock::time_point::max()) {
-      Timepoint2Timespec(context->deadline(), &deadline);
-    }
-    response->mutable_param()->set_request_deadline(deadline.tv_sec);
-  }
-}
-
-void CheckServerAuthContext(
-    const experimental::ServerContextBase* context,
-    const grpc::string& expected_transport_security_type,
-    const grpc::string& expected_client_identity) {
-  std::shared_ptr<const AuthContext> auth_ctx = context->auth_context();
-  std::vector<grpc::string_ref> tst =
-      auth_ctx->FindPropertyValues("transport_security_type");
-  EXPECT_EQ(1u, tst.size());
-  EXPECT_EQ(expected_transport_security_type, ToString(tst[0]));
-  if (expected_client_identity.empty()) {
-    EXPECT_TRUE(auth_ctx->GetPeerIdentityPropertyName().empty());
-    EXPECT_TRUE(auth_ctx->GetPeerIdentity().empty());
-    EXPECT_FALSE(auth_ctx->IsPeerAuthenticated());
-  } else {
-    auto identity = auth_ctx->GetPeerIdentity();
-    EXPECT_TRUE(auth_ctx->IsPeerAuthenticated());
-    EXPECT_EQ(1u, identity.size());
-    EXPECT_EQ(expected_client_identity, identity[0]);
-  }
-}
-
-// Returns the number of pairs in metadata that exactly match the given
-// key-value pair. Returns -1 if the pair wasn't found.
-int MetadataMatchCount(
-    const std::multimap<grpc::string_ref, grpc::string_ref>& metadata,
-    const grpc::string& key, const grpc::string& value) {
-  int count = 0;
-  for (const auto& metadatum : metadata) {
-    if (ToString(metadatum.first) == key &&
-        ToString(metadatum.second) == value) {
-      count++;
-    }
-  }
-  return count;
-}
-}  // namespace
-
-namespace {
-int GetIntValueFromMetadataHelper(
-    const char* key,
-    const std::multimap<grpc::string_ref, grpc::string_ref>& metadata,
-    int default_value) {
-  if (metadata.find(key) != metadata.end()) {
-    std::istringstream iss(ToString(metadata.find(key)->second));
-    iss >> default_value;
-    gpr_log(GPR_INFO, "%s : %d", key, default_value);
-  }
-
-  return default_value;
-}
-
-int GetIntValueFromMetadata(
-    const char* key,
-    const std::multimap<grpc::string_ref, grpc::string_ref>& metadata,
-    int default_value) {
-  return GetIntValueFromMetadataHelper(key, metadata, default_value);
-}
-
-void ServerTryCancel(ServerContext* context) {
-  EXPECT_FALSE(context->IsCancelled());
-  context->TryCancel();
-  gpr_log(GPR_INFO, "Server called TryCancel() to cancel the request");
-  // Now wait until it's really canceled
-  while (!context->IsCancelled()) {
-    gpr_sleep_until(gpr_time_add(gpr_now(GPR_CLOCK_REALTIME),
-                                 gpr_time_from_micros(1000, GPR_TIMESPAN)));
-  }
-}
-
-}  // namespace
-
-class TestMultipleServiceSignaller {
- public:
-  void ClientWaitUntilRpcStarted() {
-    std::unique_lock<std::mutex> lock(mu_);
-    cv_rpc_started_.wait(lock, [this] { return rpc_started_; });
-  }
-  void ServerWaitToContinue() {
-    std::unique_lock<std::mutex> lock(mu_);
-    cv_server_continue_.wait(lock, [this] { return server_should_continue_; });
-  }
-  void SignalClientThatRpcStarted() {
-    std::unique_lock<std::mutex> lock(mu_);
-    rpc_started_ = true;
-    cv_rpc_started_.notify_one();
-  }
-  void SignalServerToContinue() {
-    std::unique_lock<std::mutex> lock(mu_);
-    server_should_continue_ = true;
-    cv_server_continue_.notify_one();
-  }
-
- private:
-  std::mutex mu_;
-  std::condition_variable cv_rpc_started_;
-  bool rpc_started_ /* GUARDED_BY(mu_) */ = false;
-  std::condition_variable cv_server_continue_;
-  bool server_should_continue_ /* GUARDED_BY(mu_) */ = false;
-};
-
-template <typename RpcService>
-class TestMultipleServiceImpl : public RpcService {
- public:
-  TestMultipleServiceImpl() : signal_client_(false), host_() {}
-  explicit TestMultipleServiceImpl(const grpc::string& host)
-      : signal_client_(false), host_(new grpc::string(host)) {}
-
-  Status Echo(ServerContext* context, const EchoRequest* request,
-              EchoResponse* response) {
-    if (request->has_param() &&
-        request->param().server_notify_client_when_started()) {
-      signaller_.SignalClientThatRpcStarted();
-      signaller_.ServerWaitToContinue();
-    }
-
-    // A bit of sleep to make sure that short deadline tests fail
-    if (request->has_param() && request->param().server_sleep_us() > 0) {
-      gpr_sleep_until(
-          gpr_time_add(gpr_now(GPR_CLOCK_MONOTONIC),
-                       gpr_time_from_micros(request->param().server_sleep_us(),
-                                            GPR_TIMESPAN)));
-    }
-
-    if (request->has_param() && request->param().server_die()) {
-      gpr_log(GPR_ERROR, "The request should not reach application handler.");
-      GPR_ASSERT(0);
-    }
-    if (request->has_param() && request->param().has_expected_error()) {
-      const auto& error = request->param().expected_error();
-      return Status(static_cast<StatusCode>(error.code()),
-                    error.error_message(), error.binary_error_details());
-    }
-    int server_try_cancel = GetIntValueFromMetadata(
-        kServerTryCancelRequest, context->client_metadata(), DO_NOT_CANCEL);
-    if (server_try_cancel > DO_NOT_CANCEL) {
-      // Since this is a unary RPC, by the time this server handler is called,
-      // the 'request' message is already read from the client. So the scenarios
-      // in server_try_cancel don't make much sense. Just cancel the RPC as long
-      // as server_try_cancel is not DO_NOT_CANCEL
-      ServerTryCancel(context);
-      return Status::CANCELLED;
-    }
-
-    response->set_message(request->message());
-    MaybeEchoDeadline(context, request, response);
-    if (host_) {
-      response->mutable_param()->set_host(*host_);
-    }
-    if (request->has_param() && request->param().client_cancel_after_us()) {
-      {
-        std::unique_lock<std::mutex> lock(mu_);
-        signal_client_ = true;
-      }
-      while (!context->IsCancelled()) {
-        gpr_sleep_until(gpr_time_add(
-            gpr_now(GPR_CLOCK_REALTIME),
-            gpr_time_from_micros(request->param().client_cancel_after_us(),
-                                 GPR_TIMESPAN)));
-      }
-      return Status::CANCELLED;
-    } else if (request->has_param() &&
-               request->param().server_cancel_after_us()) {
-      gpr_sleep_until(gpr_time_add(
-          gpr_now(GPR_CLOCK_REALTIME),
-          gpr_time_from_micros(request->param().server_cancel_after_us(),
-                               GPR_TIMESPAN)));
-      return Status::CANCELLED;
-    } else if (!request->has_param() ||
-               !request->param().skip_cancelled_check()) {
-      EXPECT_FALSE(context->IsCancelled());
-    }
-
-    if (request->has_param() && request->param().echo_metadata_initially()) {
-      const std::multimap<grpc::string_ref, grpc::string_ref>& client_metadata =
-          context->client_metadata();
-      for (const auto& metadatum : client_metadata) {
-        context->AddInitialMetadata(ToString(metadatum.first),
-                                    ToString(metadatum.second));
-      }
-    }
-
-    if (request->has_param() && request->param().echo_metadata()) {
-      const std::multimap<grpc::string_ref, grpc::string_ref>& client_metadata =
-          context->client_metadata();
-      for (const auto& metadatum : client_metadata) {
-        context->AddTrailingMetadata(ToString(metadatum.first),
-                                     ToString(metadatum.second));
-      }
-      // Terminate rpc with error and debug info in trailer.
-      if (request->param().debug_info().stack_entries_size() ||
-          !request->param().debug_info().detail().empty()) {
-        grpc::string serialized_debug_info =
-            request->param().debug_info().SerializeAsString();
-        context->AddTrailingMetadata(kDebugInfoTrailerKey,
-                                     serialized_debug_info);
-        return Status::CANCELLED;
-      }
-    }
-    if (request->has_param() &&
-        (request->param().expected_client_identity().length() > 0 ||
-         request->param().check_auth_context())) {
-      CheckServerAuthContext(
-          context, request->param().expected_transport_security_type(),
-          request->param().expected_client_identity());
-    }
-    if (request->has_param() &&
-        request->param().response_message_length() > 0) {
-      response->set_message(
-          grpc::string(request->param().response_message_length(), '\0'));
-    }
-    if (request->has_param() && request->param().echo_peer()) {
-      response->mutable_param()->set_peer(context->peer());
-    }
-    return Status::OK;
-  }
-
-  Status Echo1(ServerContext* context, const EchoRequest* request,
-               EchoResponse* response) {
-    return Echo(context, request, response);
-  }
-
-  Status Echo2(ServerContext* context, const EchoRequest* request,
-               EchoResponse* response) {
-    return Echo(context, request, response);
-  }
-
-  Status CheckClientInitialMetadata(ServerContext* context,
-                                    const SimpleRequest* /*request*/,
-                                    SimpleResponse* /*response*/) {
-    EXPECT_EQ(MetadataMatchCount(context->client_metadata(),
-                                 kCheckClientInitialMetadataKey,
-                                 kCheckClientInitialMetadataVal),
-              1);
-    EXPECT_EQ(1u,
-              context->client_metadata().count(kCheckClientInitialMetadataKey));
-    return Status::OK;
-  }
-
-  // Unimplemented is left unimplemented to test the returned error.
-  Status RequestStream(ServerContext* context,
-                       ServerReader<EchoRequest>* reader,
-                       EchoResponse* response) {
-    // If 'server_try_cancel' is set in the metadata, the RPC is cancelled by
-    // the server by calling ServerContext::TryCancel() depending on the value:
-    //   CANCEL_BEFORE_PROCESSING: The RPC is cancelled before the server reads
-    //   any message from the client
-    //   CANCEL_DURING_PROCESSING: The RPC is cancelled while the server is
-    //   reading messages from the client
-    //   CANCEL_AFTER_PROCESSING: The RPC is cancelled after the server reads
-    //   all the messages from the client
-    int server_try_cancel = GetIntValueFromMetadata(
-        kServerTryCancelRequest, context->client_metadata(), DO_NOT_CANCEL);
-
-    EchoRequest request;
-    response->set_message("");
-
-    if (server_try_cancel == CANCEL_BEFORE_PROCESSING) {
-      ServerTryCancel(context);
-      return Status::CANCELLED;
-    }
-
-    std::thread* server_try_cancel_thd = nullptr;
-    if (server_try_cancel == CANCEL_DURING_PROCESSING) {
-      server_try_cancel_thd =
-          new std::thread([context] { ServerTryCancel(context); });
-    }
-
-    int num_msgs_read = 0;
-    while (reader->Read(&request)) {
-      response->mutable_message()->append(request.message());
-    }
-    gpr_log(GPR_INFO, "Read: %d messages", num_msgs_read);
-
-    if (server_try_cancel_thd != nullptr) {
-      server_try_cancel_thd->join();
-      delete server_try_cancel_thd;
-      return Status::CANCELLED;
-    }
-
-    if (server_try_cancel == CANCEL_AFTER_PROCESSING) {
-      ServerTryCancel(context);
-      return Status::CANCELLED;
-    }
-
-    return Status::OK;
-  }
-
-  // Return 'kNumResponseStreamMsgs' messages.
-  // TODO(yangg) make it generic by adding a parameter into EchoRequest
-  Status ResponseStream(ServerContext* context, const EchoRequest* request,
-                        ServerWriter<EchoResponse>* writer) {
-    // If server_try_cancel is set in the metadata, the RPC is cancelled by the
-    // server by calling ServerContext::TryCancel() depending on the value:
-    //   CANCEL_BEFORE_PROCESSING: The RPC is cancelled before the server writes
-    //   any messages to the client
-    //   CANCEL_DURING_PROCESSING: The RPC is cancelled while the server is
-    //   writing messages to the client
-    //   CANCEL_AFTER_PROCESSING: The RPC is cancelled after the server writes
-    //   all the messages to the client
-    int server_try_cancel = GetIntValueFromMetadata(
-        kServerTryCancelRequest, context->client_metadata(), DO_NOT_CANCEL);
-
-    int server_coalescing_api = GetIntValueFromMetadata(
-        kServerUseCoalescingApi, context->client_metadata(), 0);
-
-    int server_responses_to_send = GetIntValueFromMetadata(
-        kServerResponseStreamsToSend, context->client_metadata(),
-        kServerDefaultResponseStreamsToSend);
-
-    if (server_try_cancel == CANCEL_BEFORE_PROCESSING) {
-      ServerTryCancel(context);
-      return Status::CANCELLED;
-    }
-
-    EchoResponse response;
-    std::thread* server_try_cancel_thd = nullptr;
-    if (server_try_cancel == CANCEL_DURING_PROCESSING) {
-      server_try_cancel_thd =
-          new std::thread([context] { ServerTryCancel(context); });
-    }
-
-    for (int i = 0; i < server_responses_to_send; i++) {
-      response.set_message(request->message() + grpc::to_string(i));
-      if (i == server_responses_to_send - 1 && server_coalescing_api != 0) {
-        writer->WriteLast(response, WriteOptions());
-      } else {
-        writer->Write(response);
-      }
-    }
-
-    if (server_try_cancel_thd != nullptr) {
-      server_try_cancel_thd->join();
-      delete server_try_cancel_thd;
-      return Status::CANCELLED;
-    }
-
-    if (server_try_cancel == CANCEL_AFTER_PROCESSING) {
-      ServerTryCancel(context);
-      return Status::CANCELLED;
-    }
-
-    return Status::OK;
-  }
-
-  Status BidiStream(ServerContext* context,
-                    ServerReaderWriter<EchoResponse, EchoRequest>* stream) {
-    // If server_try_cancel is set in the metadata, the RPC is cancelled by the
-    // server by calling ServerContext::TryCancel() depending on the value:
-    //   CANCEL_BEFORE_PROCESSING: The RPC is cancelled before the server reads/
-    //   writes any messages from/to the client
-    //   CANCEL_DURING_PROCESSING: The RPC is cancelled while the server is
-    //   reading/writing messages from/to the client
-    //   CANCEL_AFTER_PROCESSING: The RPC is cancelled after the server
-    //   reads/writes all messages from/to the client
-    int server_try_cancel = GetIntValueFromMetadata(
-        kServerTryCancelRequest, context->client_metadata(), DO_NOT_CANCEL);
-
-    EchoRequest request;
-    EchoResponse response;
-
-    if (server_try_cancel == CANCEL_BEFORE_PROCESSING) {
-      ServerTryCancel(context);
-      return Status::CANCELLED;
-    }
-
-    std::thread* server_try_cancel_thd = nullptr;
-    if (server_try_cancel == CANCEL_DURING_PROCESSING) {
-      server_try_cancel_thd =
-          new std::thread([context] { ServerTryCancel(context); });
-    }
-
-    // kServerFinishAfterNReads suggests after how many reads, the server should
-    // write the last message and send status (coalesced using WriteLast)
-    int server_write_last = GetIntValueFromMetadata(
-        kServerFinishAfterNReads, context->client_metadata(), 0);
-
-    int read_counts = 0;
-    while (stream->Read(&request)) {
-      read_counts++;
-      gpr_log(GPR_INFO, "recv msg %s", request.message().c_str());
-      response.set_message(request.message());
-      if (read_counts == server_write_last) {
-        stream->WriteLast(response, WriteOptions());
-      } else {
-        stream->Write(response);
-      }
-    }
-
-    if (server_try_cancel_thd != nullptr) {
-      server_try_cancel_thd->join();
-      delete server_try_cancel_thd;
-      return Status::CANCELLED;
-    }
-
-    if (server_try_cancel == CANCEL_AFTER_PROCESSING) {
-      ServerTryCancel(context);
-      return Status::CANCELLED;
-    }
-
-    return Status::OK;
-  }
-
-  // Unimplemented is left unimplemented to test the returned error.
-  bool signal_client() {
-    std::unique_lock<std::mutex> lock(mu_);
-    return signal_client_;
-  }
-  void ClientWaitUntilRpcStarted() { signaller_.ClientWaitUntilRpcStarted(); }
-  void SignalServerToContinue() { signaller_.SignalServerToContinue(); }
-
- private:
-  bool signal_client_;
-  std::mutex mu_;
-  TestMultipleServiceSignaller signaller_;
-  std::unique_ptr<grpc::string> host_;
-};
-
-}  // namespace testing
-}  // namespace grpc
-
-#endif  // GRPC_TEST_CPP_END2END_TEST_MULTIPLE_SERVICE_IMPL_H
diff --git a/test/cpp/end2end/test_service_impl.cc b/test/cpp/end2end/test_service_impl.cc
index ad1592b..6517c5d 100644
--- a/test/cpp/end2end/test_service_impl.cc
+++ b/test/cpp/end2end/test_service_impl.cc
@@ -36,88 +36,6 @@
 namespace testing {
 namespace {
 
-// When echo_deadline is requested, deadline seen in the ServerContext is set in
-// the response in seconds.
-void MaybeEchoDeadline(experimental::ServerContextBase* context,
-                       const EchoRequest* request, EchoResponse* response) {
-  if (request->has_param() && request->param().echo_deadline()) {
-    gpr_timespec deadline = gpr_inf_future(GPR_CLOCK_REALTIME);
-    if (context->deadline() != system_clock::time_point::max()) {
-      Timepoint2Timespec(context->deadline(), &deadline);
-    }
-    response->mutable_param()->set_request_deadline(deadline.tv_sec);
-  }
-}
-
-void CheckServerAuthContext(
-    const experimental::ServerContextBase* context,
-    const grpc::string& expected_transport_security_type,
-    const grpc::string& expected_client_identity) {
-  std::shared_ptr<const AuthContext> auth_ctx = context->auth_context();
-  std::vector<grpc::string_ref> tst =
-      auth_ctx->FindPropertyValues("transport_security_type");
-  EXPECT_EQ(1u, tst.size());
-  EXPECT_EQ(expected_transport_security_type, ToString(tst[0]));
-  if (expected_client_identity.empty()) {
-    EXPECT_TRUE(auth_ctx->GetPeerIdentityPropertyName().empty());
-    EXPECT_TRUE(auth_ctx->GetPeerIdentity().empty());
-    EXPECT_FALSE(auth_ctx->IsPeerAuthenticated());
-  } else {
-    auto identity = auth_ctx->GetPeerIdentity();
-    EXPECT_TRUE(auth_ctx->IsPeerAuthenticated());
-    EXPECT_EQ(1u, identity.size());
-    EXPECT_EQ(expected_client_identity, identity[0]);
-  }
-}
-
-// Returns the number of pairs in metadata that exactly match the given
-// key-value pair. Returns -1 if the pair wasn't found.
-int MetadataMatchCount(
-    const std::multimap<grpc::string_ref, grpc::string_ref>& metadata,
-    const grpc::string& key, const grpc::string& value) {
-  int count = 0;
-  for (const auto& metadatum : metadata) {
-    if (ToString(metadatum.first) == key &&
-        ToString(metadatum.second) == value) {
-      count++;
-    }
-  }
-  return count;
-}
-}  // namespace
-
-namespace {
-int GetIntValueFromMetadataHelper(
-    const char* key,
-    const std::multimap<grpc::string_ref, grpc::string_ref>& metadata,
-    int default_value) {
-  if (metadata.find(key) != metadata.end()) {
-    std::istringstream iss(ToString(metadata.find(key)->second));
-    iss >> default_value;
-    gpr_log(GPR_INFO, "%s : %d", key, default_value);
-  }
-
-  return default_value;
-}
-
-int GetIntValueFromMetadata(
-    const char* key,
-    const std::multimap<grpc::string_ref, grpc::string_ref>& metadata,
-    int default_value) {
-  return GetIntValueFromMetadataHelper(key, metadata, default_value);
-}
-
-void ServerTryCancel(ServerContext* context) {
-  EXPECT_FALSE(context->IsCancelled());
-  context->TryCancel();
-  gpr_log(GPR_INFO, "Server called TryCancel() to cancel the request");
-  // Now wait until it's really canceled
-  while (!context->IsCancelled()) {
-    gpr_sleep_until(gpr_time_add(gpr_now(GPR_CLOCK_REALTIME),
-                                 gpr_time_from_micros(1000, GPR_TIMESPAN)));
-  }
-}
-
 void ServerTryCancelNonblocking(experimental::CallbackServerContext* context) {
   EXPECT_FALSE(context->IsCancelled());
   context->TryCancel();
@@ -127,304 +45,6 @@
 
 }  // namespace
 
-Status TestServiceImpl::Echo(ServerContext* context, const EchoRequest* request,
-                             EchoResponse* response) {
-  if (request->has_param() &&
-      request->param().server_notify_client_when_started()) {
-    signaller_.SignalClientThatRpcStarted();
-    signaller_.ServerWaitToContinue();
-  }
-
-  // A bit of sleep to make sure that short deadline tests fail
-  if (request->has_param() && request->param().server_sleep_us() > 0) {
-    gpr_sleep_until(
-        gpr_time_add(gpr_now(GPR_CLOCK_MONOTONIC),
-                     gpr_time_from_micros(request->param().server_sleep_us(),
-                                          GPR_TIMESPAN)));
-  }
-
-  if (request->has_param() && request->param().server_die()) {
-    gpr_log(GPR_ERROR, "The request should not reach application handler.");
-    GPR_ASSERT(0);
-  }
-  if (request->has_param() && request->param().has_expected_error()) {
-    const auto& error = request->param().expected_error();
-    return Status(static_cast<StatusCode>(error.code()), error.error_message(),
-                  error.binary_error_details());
-  }
-  int server_try_cancel = GetIntValueFromMetadata(
-      kServerTryCancelRequest, context->client_metadata(), DO_NOT_CANCEL);
-  if (server_try_cancel > DO_NOT_CANCEL) {
-    // Since this is a unary RPC, by the time this server handler is called,
-    // the 'request' message is already read from the client. So the scenarios
-    // in server_try_cancel don't make much sense. Just cancel the RPC as long
-    // as server_try_cancel is not DO_NOT_CANCEL
-    ServerTryCancel(context);
-    return Status::CANCELLED;
-  }
-
-  response->set_message(request->message());
-  MaybeEchoDeadline(context, request, response);
-  if (host_) {
-    response->mutable_param()->set_host(*host_);
-  }
-  if (request->has_param() && request->param().client_cancel_after_us()) {
-    {
-      std::unique_lock<std::mutex> lock(mu_);
-      signal_client_ = true;
-    }
-    while (!context->IsCancelled()) {
-      gpr_sleep_until(gpr_time_add(
-          gpr_now(GPR_CLOCK_REALTIME),
-          gpr_time_from_micros(request->param().client_cancel_after_us(),
-                               GPR_TIMESPAN)));
-    }
-    return Status::CANCELLED;
-  } else if (request->has_param() &&
-             request->param().server_cancel_after_us()) {
-    gpr_sleep_until(gpr_time_add(
-        gpr_now(GPR_CLOCK_REALTIME),
-        gpr_time_from_micros(request->param().server_cancel_after_us(),
-                             GPR_TIMESPAN)));
-    return Status::CANCELLED;
-  } else if (!request->has_param() ||
-             !request->param().skip_cancelled_check()) {
-    EXPECT_FALSE(context->IsCancelled());
-  }
-
-  if (request->has_param() && request->param().echo_metadata_initially()) {
-    const std::multimap<grpc::string_ref, grpc::string_ref>& client_metadata =
-        context->client_metadata();
-    for (const auto& metadatum : client_metadata) {
-      context->AddInitialMetadata(ToString(metadatum.first),
-                                  ToString(metadatum.second));
-    }
-  }
-
-  if (request->has_param() && request->param().echo_metadata()) {
-    const std::multimap<grpc::string_ref, grpc::string_ref>& client_metadata =
-        context->client_metadata();
-    for (const auto& metadatum : client_metadata) {
-      context->AddTrailingMetadata(ToString(metadatum.first),
-                                   ToString(metadatum.second));
-    }
-    // Terminate rpc with error and debug info in trailer.
-    if (request->param().debug_info().stack_entries_size() ||
-        !request->param().debug_info().detail().empty()) {
-      grpc::string serialized_debug_info =
-          request->param().debug_info().SerializeAsString();
-      context->AddTrailingMetadata(kDebugInfoTrailerKey, serialized_debug_info);
-      return Status::CANCELLED;
-    }
-  }
-  if (request->has_param() &&
-      (request->param().expected_client_identity().length() > 0 ||
-       request->param().check_auth_context())) {
-    CheckServerAuthContext(context,
-                           request->param().expected_transport_security_type(),
-                           request->param().expected_client_identity());
-  }
-  if (request->has_param() && request->param().response_message_length() > 0) {
-    response->set_message(
-        grpc::string(request->param().response_message_length(), '\0'));
-  }
-  if (request->has_param() && request->param().echo_peer()) {
-    response->mutable_param()->set_peer(context->peer());
-  }
-  return Status::OK;
-}
-
-Status TestServiceImpl::Echo1(ServerContext* context,
-                              const EchoRequest* request,
-                              EchoResponse* response) {
-  return Echo(context, request, response);
-}
-
-Status TestServiceImpl::Echo2(ServerContext* context,
-                              const EchoRequest* request,
-                              EchoResponse* response) {
-  return Echo(context, request, response);
-}
-
-Status TestServiceImpl::CheckClientInitialMetadata(
-    ServerContext* context, const SimpleRequest* /*request*/,
-    SimpleResponse* /*response*/) {
-  EXPECT_EQ(MetadataMatchCount(context->client_metadata(),
-                               kCheckClientInitialMetadataKey,
-                               kCheckClientInitialMetadataVal),
-            1);
-  EXPECT_EQ(1u,
-            context->client_metadata().count(kCheckClientInitialMetadataKey));
-  return Status::OK;
-}
-
-// Unimplemented is left unimplemented to test the returned error.
-
-Status TestServiceImpl::RequestStream(ServerContext* context,
-                                      ServerReader<EchoRequest>* reader,
-                                      EchoResponse* response) {
-  // If 'server_try_cancel' is set in the metadata, the RPC is cancelled by
-  // the server by calling ServerContext::TryCancel() depending on the value:
-  //   CANCEL_BEFORE_PROCESSING: The RPC is cancelled before the server reads
-  //   any message from the client
-  //   CANCEL_DURING_PROCESSING: The RPC is cancelled while the server is
-  //   reading messages from the client
-  //   CANCEL_AFTER_PROCESSING: The RPC is cancelled after the server reads
-  //   all the messages from the client
-  int server_try_cancel = GetIntValueFromMetadata(
-      kServerTryCancelRequest, context->client_metadata(), DO_NOT_CANCEL);
-
-  EchoRequest request;
-  response->set_message("");
-
-  if (server_try_cancel == CANCEL_BEFORE_PROCESSING) {
-    ServerTryCancel(context);
-    return Status::CANCELLED;
-  }
-
-  std::thread* server_try_cancel_thd = nullptr;
-  if (server_try_cancel == CANCEL_DURING_PROCESSING) {
-    server_try_cancel_thd =
-        new std::thread([context] { ServerTryCancel(context); });
-  }
-
-  int num_msgs_read = 0;
-  while (reader->Read(&request)) {
-    response->mutable_message()->append(request.message());
-  }
-  gpr_log(GPR_INFO, "Read: %d messages", num_msgs_read);
-
-  if (server_try_cancel_thd != nullptr) {
-    server_try_cancel_thd->join();
-    delete server_try_cancel_thd;
-    return Status::CANCELLED;
-  }
-
-  if (server_try_cancel == CANCEL_AFTER_PROCESSING) {
-    ServerTryCancel(context);
-    return Status::CANCELLED;
-  }
-
-  return Status::OK;
-}
-
-// Return 'kNumResponseStreamMsgs' messages.
-// TODO(yangg) make it generic by adding a parameter into EchoRequest
-Status TestServiceImpl::ResponseStream(ServerContext* context,
-                                       const EchoRequest* request,
-                                       ServerWriter<EchoResponse>* writer) {
-  // If server_try_cancel is set in the metadata, the RPC is cancelled by the
-  // server by calling ServerContext::TryCancel() depending on the value:
-  //   CANCEL_BEFORE_PROCESSING: The RPC is cancelled before the server writes
-  //   any messages to the client
-  //   CANCEL_DURING_PROCESSING: The RPC is cancelled while the server is
-  //   writing messages to the client
-  //   CANCEL_AFTER_PROCESSING: The RPC is cancelled after the server writes
-  //   all the messages to the client
-  int server_try_cancel = GetIntValueFromMetadata(
-      kServerTryCancelRequest, context->client_metadata(), DO_NOT_CANCEL);
-
-  int server_coalescing_api = GetIntValueFromMetadata(
-      kServerUseCoalescingApi, context->client_metadata(), 0);
-
-  int server_responses_to_send = GetIntValueFromMetadata(
-      kServerResponseStreamsToSend, context->client_metadata(),
-      kServerDefaultResponseStreamsToSend);
-
-  if (server_try_cancel == CANCEL_BEFORE_PROCESSING) {
-    ServerTryCancel(context);
-    return Status::CANCELLED;
-  }
-
-  EchoResponse response;
-  std::thread* server_try_cancel_thd = nullptr;
-  if (server_try_cancel == CANCEL_DURING_PROCESSING) {
-    server_try_cancel_thd =
-        new std::thread([context] { ServerTryCancel(context); });
-  }
-
-  for (int i = 0; i < server_responses_to_send; i++) {
-    response.set_message(request->message() + grpc::to_string(i));
-    if (i == server_responses_to_send - 1 && server_coalescing_api != 0) {
-      writer->WriteLast(response, WriteOptions());
-    } else {
-      writer->Write(response);
-    }
-  }
-
-  if (server_try_cancel_thd != nullptr) {
-    server_try_cancel_thd->join();
-    delete server_try_cancel_thd;
-    return Status::CANCELLED;
-  }
-
-  if (server_try_cancel == CANCEL_AFTER_PROCESSING) {
-    ServerTryCancel(context);
-    return Status::CANCELLED;
-  }
-
-  return Status::OK;
-}
-
-Status TestServiceImpl::BidiStream(
-    ServerContext* context,
-    ServerReaderWriter<EchoResponse, EchoRequest>* stream) {
-  // If server_try_cancel is set in the metadata, the RPC is cancelled by the
-  // server by calling ServerContext::TryCancel() depending on the value:
-  //   CANCEL_BEFORE_PROCESSING: The RPC is cancelled before the server reads/
-  //   writes any messages from/to the client
-  //   CANCEL_DURING_PROCESSING: The RPC is cancelled while the server is
-  //   reading/writing messages from/to the client
-  //   CANCEL_AFTER_PROCESSING: The RPC is cancelled after the server
-  //   reads/writes all messages from/to the client
-  int server_try_cancel = GetIntValueFromMetadata(
-      kServerTryCancelRequest, context->client_metadata(), DO_NOT_CANCEL);
-
-  EchoRequest request;
-  EchoResponse response;
-
-  if (server_try_cancel == CANCEL_BEFORE_PROCESSING) {
-    ServerTryCancel(context);
-    return Status::CANCELLED;
-  }
-
-  std::thread* server_try_cancel_thd = nullptr;
-  if (server_try_cancel == CANCEL_DURING_PROCESSING) {
-    server_try_cancel_thd =
-        new std::thread([context] { ServerTryCancel(context); });
-  }
-
-  // kServerFinishAfterNReads suggests after how many reads, the server should
-  // write the last message and send status (coalesced using WriteLast)
-  int server_write_last = GetIntValueFromMetadata(
-      kServerFinishAfterNReads, context->client_metadata(), 0);
-
-  int read_counts = 0;
-  while (stream->Read(&request)) {
-    read_counts++;
-    gpr_log(GPR_INFO, "recv msg %s", request.message().c_str());
-    response.set_message(request.message());
-    if (read_counts == server_write_last) {
-      stream->WriteLast(response, WriteOptions());
-    } else {
-      stream->Write(response);
-    }
-  }
-
-  if (server_try_cancel_thd != nullptr) {
-    server_try_cancel_thd->join();
-    delete server_try_cancel_thd;
-    return Status::CANCELLED;
-  }
-
-  if (server_try_cancel == CANCEL_AFTER_PROCESSING) {
-    ServerTryCancel(context);
-    return Status::CANCELLED;
-  }
-
-  return Status::OK;
-}
-
 experimental::ServerUnaryReactor* CallbackTestServiceImpl::Echo(
     experimental::CallbackServerContext* context, const EchoRequest* request,
     EchoResponse* response) {
diff --git a/test/cpp/end2end/test_service_impl.h b/test/cpp/end2end/test_service_impl.h
index e41359f..905f2cb 100644
--- a/test/cpp/end2end/test_service_impl.h
+++ b/test/cpp/end2end/test_service_impl.h
@@ -15,6 +15,7 @@
  * limitations under the License.
  *
  */
+
 #ifndef GRPC_TEST_CPP_END2END_TEST_SERVICE_IMPL_H
 #define GRPC_TEST_CPP_END2END_TEST_SERVICE_IMPL_H
 
@@ -23,9 +24,19 @@
 #include <mutex>
 
 #include <grpc/grpc.h>
+#include <grpc/support/log.h>
+#include <grpcpp/alarm.h>
+#include <grpcpp/security/credentials.h>
 #include <grpcpp/server_context.h>
+#include <gtest/gtest.h>
+
+#include <string>
+#include <thread>
 
 #include "src/proto/grpc/testing/echo.grpc.pb.h"
+#include "test/cpp/util/string_ref_helper.h"
+
+using std::chrono::system_clock;
 
 namespace grpc {
 namespace testing {
@@ -46,6 +57,92 @@
   CANCEL_AFTER_PROCESSING
 } ServerTryCancelRequestPhase;
 
+namespace {
+
+// When echo_deadline is requested, deadline seen in the ServerContext is set in
+// the response in seconds.
+void MaybeEchoDeadline(experimental::ServerContextBase* context,
+                       const EchoRequest* request, EchoResponse* response) {
+  if (request->has_param() && request->param().echo_deadline()) {
+    gpr_timespec deadline = gpr_inf_future(GPR_CLOCK_REALTIME);
+    if (context->deadline() != system_clock::time_point::max()) {
+      Timepoint2Timespec(context->deadline(), &deadline);
+    }
+    response->mutable_param()->set_request_deadline(deadline.tv_sec);
+  }
+}
+
+void CheckServerAuthContext(
+    const experimental::ServerContextBase* context,
+    const grpc::string& expected_transport_security_type,
+    const grpc::string& expected_client_identity) {
+  std::shared_ptr<const AuthContext> auth_ctx = context->auth_context();
+  std::vector<grpc::string_ref> tst =
+      auth_ctx->FindPropertyValues("transport_security_type");
+  EXPECT_EQ(1u, tst.size());
+  EXPECT_EQ(expected_transport_security_type, ToString(tst[0]));
+  if (expected_client_identity.empty()) {
+    EXPECT_TRUE(auth_ctx->GetPeerIdentityPropertyName().empty());
+    EXPECT_TRUE(auth_ctx->GetPeerIdentity().empty());
+    EXPECT_FALSE(auth_ctx->IsPeerAuthenticated());
+  } else {
+    auto identity = auth_ctx->GetPeerIdentity();
+    EXPECT_TRUE(auth_ctx->IsPeerAuthenticated());
+    EXPECT_EQ(1u, identity.size());
+    EXPECT_EQ(expected_client_identity, identity[0]);
+  }
+}
+
+// Returns the number of pairs in metadata that exactly match the given
+// key-value pair. Returns -1 if the pair wasn't found.
+int MetadataMatchCount(
+    const std::multimap<grpc::string_ref, grpc::string_ref>& metadata,
+    const grpc::string& key, const grpc::string& value) {
+  int count = 0;
+  for (const auto& metadatum : metadata) {
+    if (ToString(metadatum.first) == key &&
+        ToString(metadatum.second) == value) {
+      count++;
+    }
+  }
+  return count;
+}
+}  // namespace
+
+namespace {
+int GetIntValueFromMetadataHelper(
+    const char* key,
+    const std::multimap<grpc::string_ref, grpc::string_ref>& metadata,
+    int default_value) {
+  if (metadata.find(key) != metadata.end()) {
+    std::istringstream iss(ToString(metadata.find(key)->second));
+    iss >> default_value;
+    gpr_log(GPR_INFO, "%s : %d", key, default_value);
+  }
+
+  return default_value;
+}
+
+int GetIntValueFromMetadata(
+    const char* key,
+    const std::multimap<grpc::string_ref, grpc::string_ref>& metadata,
+    int default_value) {
+  return GetIntValueFromMetadataHelper(key, metadata, default_value);
+}
+
+void ServerTryCancel(ServerContext* context) {
+  EXPECT_FALSE(context->IsCancelled());
+  context->TryCancel();
+  gpr_log(GPR_INFO, "Server called TryCancel() to cancel the request");
+  // Now wait until it's really canceled
+  while (!context->IsCancelled()) {
+    gpr_sleep_until(gpr_time_add(gpr_now(GPR_CLOCK_REALTIME),
+                                 gpr_time_from_micros(1000, GPR_TIMESPAN)));
+  }
+}
+
+}  // namespace
+
 class TestServiceSignaller {
  public:
   void ClientWaitUntilRpcStarted() {
@@ -75,38 +172,309 @@
   bool server_should_continue_ /* GUARDED_BY(mu_) */ = false;
 };
 
-class TestServiceImpl : public ::grpc::testing::EchoTestService::Service {
+template <typename RpcService>
+class TestMultipleServiceImpl : public RpcService {
  public:
-  TestServiceImpl() : signal_client_(false), host_() {}
-  explicit TestServiceImpl(const grpc::string& host)
+  TestMultipleServiceImpl() : signal_client_(false), host_() {}
+  explicit TestMultipleServiceImpl(const grpc::string& host)
       : signal_client_(false), host_(new grpc::string(host)) {}
 
   Status Echo(ServerContext* context, const EchoRequest* request,
-              EchoResponse* response) override;
+              EchoResponse* response) {
+    if (request->has_param() &&
+        request->param().server_notify_client_when_started()) {
+      signaller_.SignalClientThatRpcStarted();
+      signaller_.ServerWaitToContinue();
+    }
+
+    // A bit of sleep to make sure that short deadline tests fail
+    if (request->has_param() && request->param().server_sleep_us() > 0) {
+      gpr_sleep_until(
+          gpr_time_add(gpr_now(GPR_CLOCK_MONOTONIC),
+                       gpr_time_from_micros(request->param().server_sleep_us(),
+                                            GPR_TIMESPAN)));
+    }
+
+    if (request->has_param() && request->param().server_die()) {
+      gpr_log(GPR_ERROR, "The request should not reach application handler.");
+      GPR_ASSERT(0);
+    }
+    if (request->has_param() && request->param().has_expected_error()) {
+      const auto& error = request->param().expected_error();
+      return Status(static_cast<StatusCode>(error.code()),
+                    error.error_message(), error.binary_error_details());
+    }
+    int server_try_cancel = GetIntValueFromMetadata(
+        kServerTryCancelRequest, context->client_metadata(), DO_NOT_CANCEL);
+    if (server_try_cancel > DO_NOT_CANCEL) {
+      // Since this is a unary RPC, by the time this server handler is called,
+      // the 'request' message is already read from the client. So the scenarios
+      // in server_try_cancel don't make much sense. Just cancel the RPC as long
+      // as server_try_cancel is not DO_NOT_CANCEL
+      ServerTryCancel(context);
+      return Status::CANCELLED;
+    }
+
+    response->set_message(request->message());
+    MaybeEchoDeadline(context, request, response);
+    if (host_) {
+      response->mutable_param()->set_host(*host_);
+    }
+    if (request->has_param() && request->param().client_cancel_after_us()) {
+      {
+        std::unique_lock<std::mutex> lock(mu_);
+        signal_client_ = true;
+      }
+      while (!context->IsCancelled()) {
+        gpr_sleep_until(gpr_time_add(
+            gpr_now(GPR_CLOCK_REALTIME),
+            gpr_time_from_micros(request->param().client_cancel_after_us(),
+                                 GPR_TIMESPAN)));
+      }
+      return Status::CANCELLED;
+    } else if (request->has_param() &&
+               request->param().server_cancel_after_us()) {
+      gpr_sleep_until(gpr_time_add(
+          gpr_now(GPR_CLOCK_REALTIME),
+          gpr_time_from_micros(request->param().server_cancel_after_us(),
+                               GPR_TIMESPAN)));
+      return Status::CANCELLED;
+    } else if (!request->has_param() ||
+               !request->param().skip_cancelled_check()) {
+      EXPECT_FALSE(context->IsCancelled());
+    }
+
+    if (request->has_param() && request->param().echo_metadata_initially()) {
+      const std::multimap<grpc::string_ref, grpc::string_ref>& client_metadata =
+          context->client_metadata();
+      for (const auto& metadatum : client_metadata) {
+        context->AddInitialMetadata(ToString(metadatum.first),
+                                    ToString(metadatum.second));
+      }
+    }
+
+    if (request->has_param() && request->param().echo_metadata()) {
+      const std::multimap<grpc::string_ref, grpc::string_ref>& client_metadata =
+          context->client_metadata();
+      for (const auto& metadatum : client_metadata) {
+        context->AddTrailingMetadata(ToString(metadatum.first),
+                                     ToString(metadatum.second));
+      }
+      // Terminate rpc with error and debug info in trailer.
+      if (request->param().debug_info().stack_entries_size() ||
+          !request->param().debug_info().detail().empty()) {
+        grpc::string serialized_debug_info =
+            request->param().debug_info().SerializeAsString();
+        context->AddTrailingMetadata(kDebugInfoTrailerKey,
+                                     serialized_debug_info);
+        return Status::CANCELLED;
+      }
+    }
+    if (request->has_param() &&
+        (request->param().expected_client_identity().length() > 0 ||
+         request->param().check_auth_context())) {
+      CheckServerAuthContext(
+          context, request->param().expected_transport_security_type(),
+          request->param().expected_client_identity());
+    }
+    if (request->has_param() &&
+        request->param().response_message_length() > 0) {
+      response->set_message(
+          grpc::string(request->param().response_message_length(), '\0'));
+    }
+    if (request->has_param() && request->param().echo_peer()) {
+      response->mutable_param()->set_peer(context->peer());
+    }
+    return Status::OK;
+  }
 
   Status Echo1(ServerContext* context, const EchoRequest* request,
-               EchoResponse* response) override;
+               EchoResponse* response) {
+    return Echo(context, request, response);
+  }
 
   Status Echo2(ServerContext* context, const EchoRequest* request,
-               EchoResponse* response) override;
+               EchoResponse* response) {
+    return Echo(context, request, response);
+  }
 
   Status CheckClientInitialMetadata(ServerContext* context,
-                                    const SimpleRequest* request,
-                                    SimpleResponse* response) override;
+                                    const SimpleRequest* /*request*/,
+                                    SimpleResponse* /*response*/) {
+    EXPECT_EQ(MetadataMatchCount(context->client_metadata(),
+                                 kCheckClientInitialMetadataKey,
+                                 kCheckClientInitialMetadataVal),
+              1);
+    EXPECT_EQ(1u,
+              context->client_metadata().count(kCheckClientInitialMetadataKey));
+    return Status::OK;
+  }
 
   // Unimplemented is left unimplemented to test the returned error.
-
   Status RequestStream(ServerContext* context,
                        ServerReader<EchoRequest>* reader,
-                       EchoResponse* response) override;
+                       EchoResponse* response) {
+    // If 'server_try_cancel' is set in the metadata, the RPC is cancelled by
+    // the server by calling ServerContext::TryCancel() depending on the value:
+    //   CANCEL_BEFORE_PROCESSING: The RPC is cancelled before the server reads
+    //   any message from the client
+    //   CANCEL_DURING_PROCESSING: The RPC is cancelled while the server is
+    //   reading messages from the client
+    //   CANCEL_AFTER_PROCESSING: The RPC is cancelled after the server reads
+    //   all the messages from the client
+    int server_try_cancel = GetIntValueFromMetadata(
+        kServerTryCancelRequest, context->client_metadata(), DO_NOT_CANCEL);
 
+    EchoRequest request;
+    response->set_message("");
+
+    if (server_try_cancel == CANCEL_BEFORE_PROCESSING) {
+      ServerTryCancel(context);
+      return Status::CANCELLED;
+    }
+
+    std::thread* server_try_cancel_thd = nullptr;
+    if (server_try_cancel == CANCEL_DURING_PROCESSING) {
+      server_try_cancel_thd =
+          new std::thread([context] { ServerTryCancel(context); });
+    }
+
+    int num_msgs_read = 0;
+    while (reader->Read(&request)) {
+      response->mutable_message()->append(request.message());
+    }
+    gpr_log(GPR_INFO, "Read: %d messages", num_msgs_read);
+
+    if (server_try_cancel_thd != nullptr) {
+      server_try_cancel_thd->join();
+      delete server_try_cancel_thd;
+      return Status::CANCELLED;
+    }
+
+    if (server_try_cancel == CANCEL_AFTER_PROCESSING) {
+      ServerTryCancel(context);
+      return Status::CANCELLED;
+    }
+
+    return Status::OK;
+  }
+
+  // Return 'kNumResponseStreamMsgs' messages.
+  // TODO(yangg) make it generic by adding a parameter into EchoRequest
   Status ResponseStream(ServerContext* context, const EchoRequest* request,
-                        ServerWriter<EchoResponse>* writer) override;
+                        ServerWriter<EchoResponse>* writer) {
+    // If server_try_cancel is set in the metadata, the RPC is cancelled by the
+    // server by calling ServerContext::TryCancel() depending on the value:
+    //   CANCEL_BEFORE_PROCESSING: The RPC is cancelled before the server writes
+    //   any messages to the client
+    //   CANCEL_DURING_PROCESSING: The RPC is cancelled while the server is
+    //   writing messages to the client
+    //   CANCEL_AFTER_PROCESSING: The RPC is cancelled after the server writes
+    //   all the messages to the client
+    int server_try_cancel = GetIntValueFromMetadata(
+        kServerTryCancelRequest, context->client_metadata(), DO_NOT_CANCEL);
 
-  Status BidiStream(
-      ServerContext* context,
-      ServerReaderWriter<EchoResponse, EchoRequest>* stream) override;
+    int server_coalescing_api = GetIntValueFromMetadata(
+        kServerUseCoalescingApi, context->client_metadata(), 0);
 
+    int server_responses_to_send = GetIntValueFromMetadata(
+        kServerResponseStreamsToSend, context->client_metadata(),
+        kServerDefaultResponseStreamsToSend);
+
+    if (server_try_cancel == CANCEL_BEFORE_PROCESSING) {
+      ServerTryCancel(context);
+      return Status::CANCELLED;
+    }
+
+    EchoResponse response;
+    std::thread* server_try_cancel_thd = nullptr;
+    if (server_try_cancel == CANCEL_DURING_PROCESSING) {
+      server_try_cancel_thd =
+          new std::thread([context] { ServerTryCancel(context); });
+    }
+
+    for (int i = 0; i < server_responses_to_send; i++) {
+      response.set_message(request->message() + grpc::to_string(i));
+      if (i == server_responses_to_send - 1 && server_coalescing_api != 0) {
+        writer->WriteLast(response, WriteOptions());
+      } else {
+        writer->Write(response);
+      }
+    }
+
+    if (server_try_cancel_thd != nullptr) {
+      server_try_cancel_thd->join();
+      delete server_try_cancel_thd;
+      return Status::CANCELLED;
+    }
+
+    if (server_try_cancel == CANCEL_AFTER_PROCESSING) {
+      ServerTryCancel(context);
+      return Status::CANCELLED;
+    }
+
+    return Status::OK;
+  }
+
+  Status BidiStream(ServerContext* context,
+                    ServerReaderWriter<EchoResponse, EchoRequest>* stream) {
+    // If server_try_cancel is set in the metadata, the RPC is cancelled by the
+    // server by calling ServerContext::TryCancel() depending on the value:
+    //   CANCEL_BEFORE_PROCESSING: The RPC is cancelled before the server reads/
+    //   writes any messages from/to the client
+    //   CANCEL_DURING_PROCESSING: The RPC is cancelled while the server is
+    //   reading/writing messages from/to the client
+    //   CANCEL_AFTER_PROCESSING: The RPC is cancelled after the server
+    //   reads/writes all messages from/to the client
+    int server_try_cancel = GetIntValueFromMetadata(
+        kServerTryCancelRequest, context->client_metadata(), DO_NOT_CANCEL);
+
+    EchoRequest request;
+    EchoResponse response;
+
+    if (server_try_cancel == CANCEL_BEFORE_PROCESSING) {
+      ServerTryCancel(context);
+      return Status::CANCELLED;
+    }
+
+    std::thread* server_try_cancel_thd = nullptr;
+    if (server_try_cancel == CANCEL_DURING_PROCESSING) {
+      server_try_cancel_thd =
+          new std::thread([context] { ServerTryCancel(context); });
+    }
+
+    // kServerFinishAfterNReads suggests after how many reads, the server should
+    // write the last message and send status (coalesced using WriteLast)
+    int server_write_last = GetIntValueFromMetadata(
+        kServerFinishAfterNReads, context->client_metadata(), 0);
+
+    int read_counts = 0;
+    while (stream->Read(&request)) {
+      read_counts++;
+      gpr_log(GPR_INFO, "recv msg %s", request.message().c_str());
+      response.set_message(request.message());
+      if (read_counts == server_write_last) {
+        stream->WriteLast(response, WriteOptions());
+      } else {
+        stream->Write(response);
+      }
+    }
+
+    if (server_try_cancel_thd != nullptr) {
+      server_try_cancel_thd->join();
+      delete server_try_cancel_thd;
+      return Status::CANCELLED;
+    }
+
+    if (server_try_cancel == CANCEL_AFTER_PROCESSING) {
+      ServerTryCancel(context);
+      return Status::CANCELLED;
+    }
+
+    return Status::OK;
+  }
+
+  // Unimplemented is left unimplemented to test the returned error.
   bool signal_client() {
     std::unique_lock<std::mutex> lock(mu_);
     return signal_client_;
@@ -162,6 +530,9 @@
   std::unique_ptr<grpc::string> host_;
 };
 
+using TestServiceImpl =
+    TestMultipleServiceImpl<::grpc::testing::EchoTestService::Service>;
+
 }  // namespace testing
 }  // namespace grpc
 
diff --git a/test/cpp/end2end/xds_end2end_test.cc b/test/cpp/end2end/xds_end2end_test.cc
index 2665f9e..2e9d2ae 100644
--- a/test/cpp/end2end/xds_end2end_test.cc
+++ b/test/cpp/end2end/xds_end2end_test.cc
@@ -57,7 +57,7 @@
 
 #include "test/core/util/port.h"
 #include "test/core/util/test_config.h"
-#include "test/cpp/end2end/test_multiple_service_impl.h"
+#include "test/cpp/end2end/test_service_impl.h"
 
 #include "src/proto/grpc/testing/echo.grpc.pb.h"
 #include "src/proto/grpc/testing/xds/ads_for_test.grpc.pb.h"
@@ -2316,6 +2316,7 @@
   Listener listener =
       balancers_[0]->ads_service()->BuildListener(new_route_config);
   balancers_[0]->ads_service()->SetLdsResource(listener, kDefaultResourceName);
+  WaitForAllBackends(0, 2);
   CheckRpcSendOk(kNumRpcs, 1000, true);
   CheckEcho1RpcSendOk(kNumRpcs, 1000, true);
   CheckEcho2RpcSendOk(kNumRpcs, 1000, true);