HBASE-25071 ReplicationServer support start ReplicationSource internal

Author: infraio

hbase-protocol-shaded/src/main/protobuf/server/replication/ReplicationServer.proto

@@ -24,9 +24,21 @@ option java_generic_services = true;
 option java_generate_equals_and_hash = true;
 option optimize_for = SPEED;
 
+import "HBase.proto";
 import "server/region/Admin.proto";
 
+message StartReplicationSourceRequest {
+  required ServerName server_name = 1;
+  required string queue_id = 2;
+}
+
+message StartReplicationSourceResponse {
+}
+
 service ReplicationServerService {
   rpc ReplicateWALEntry(ReplicateWALEntryRequest)
     returns(ReplicateWALEntryResponse);
-}
+
+  rpc StartReplicationSource(StartReplicationSourceRequest)
+    returns(StartReplicationSourceResponse);
+}

hbase-server/src/main/java/org/apache/hadoop/hbase/master/MasterRpcServices.java

@@ -3396,7 +3396,7 @@ public ListReplicationSinkServersResponse listReplicationSinkServers(
       if (master.getMasterCoprocessorHost() != null) {
         master.getMasterCoprocessorHost().preListReplicationSinkServers();
       }
-      builder.addAllServerName(master.listReplicationSinkServers().stream()
+      builder.addAllServerName(master.getReplicationServerManager().getOnlineServersList().stream()
         .map(ProtobufUtil::toServerName).collect(Collectors.toList()));
       if (master.getMasterCoprocessorHost() != null) {
         master.getMasterCoprocessorHost().postListReplicationSinkServers();

hbase-server/src/main/java/org/apache/hadoop/hbase/regionserver/RSRpcServices.java

@@ -257,7 +257,6 @@
 import org.apache.hadoop.hbase.shaded.protobuf.generated.QuotaProtos.GetSpaceQuotaSnapshotsResponse;
 import org.apache.hadoop.hbase.shaded.protobuf.generated.QuotaProtos.GetSpaceQuotaSnapshotsResponse.TableQuotaSnapshot;
 import org.apache.hadoop.hbase.shaded.protobuf.generated.RPCProtos.RequestHeader;
-import org.apache.hadoop.hbase.shaded.protobuf.generated.ReplicationServerProtos.ReplicationServerService;
 import org.apache.hadoop.hbase.shaded.protobuf.generated.TooSlowLog.SlowLogPayload;
 import org.apache.hadoop.hbase.shaded.protobuf.generated.WALProtos.BulkLoadDescriptor;
 import org.apache.hadoop.hbase.shaded.protobuf.generated.WALProtos.CompactionDescriptor;
@@ -271,7 +270,7 @@
 @SuppressWarnings("deprecation")
 public class RSRpcServices implements HBaseRPCErrorHandler,
     AdminService.BlockingInterface, ClientService.BlockingInterface, PriorityFunction,
-    ConfigurationObserver, ReplicationServerService.BlockingInterface {
+    ConfigurationObserver {
   protected static final Logger LOG = LoggerFactory.getLogger(RSRpcServices.class);
 
   /** RPC scheduler to use for the region server. */
@@ -1491,9 +1490,6 @@ protected List<BlockingServiceAndInterface> getServices() {
       bssi.add(new BlockingServiceAndInterface(
           AdminService.newReflectiveBlockingService(this),
           AdminService.BlockingInterface.class));
-      bssi.add(new BlockingServiceAndInterface(
-          ReplicationServerService.newReflectiveBlockingService(this),
-          ReplicationServerService.BlockingInterface.class));
     }
     return new org.apache.hbase.thirdparty.com.google.common.collect.
         ImmutableList.Builder<BlockingServiceAndInterface>().addAll(bssi).build();

hbase-server/src/main/java/org/apache/hadoop/hbase/replication/HBaseReplicationEndpoint.java

@@ -42,11 +42,15 @@
 import org.apache.hadoop.hbase.client.AsyncReplicationServerAdmin;
 import org.apache.hadoop.hbase.client.ClusterConnectionFactory;
 import org.apache.hadoop.hbase.protobuf.ReplicationProtobufUtil;
+import org.apache.hadoop.hbase.ScheduledChore;
+import org.apache.hadoop.hbase.Server;
 import org.apache.hadoop.hbase.security.User;
-import org.apache.hadoop.hbase.util.FutureUtils;
 import org.apache.hadoop.hbase.wal.WAL;
-import org.apache.hadoop.hbase.zookeeper.ZKClusterId;
 import org.apache.hadoop.hbase.zookeeper.ZKListener;
+import org.apache.hadoop.hbase.ServerName;
+import org.apache.hadoop.hbase.security.UserProvider;
+import org.apache.hadoop.hbase.util.FutureUtils;
+import org.apache.hadoop.hbase.zookeeper.ZKClusterId;
 import org.apache.hadoop.hbase.zookeeper.ZKUtil;
 import org.apache.hadoop.hbase.zookeeper.ZKWatcher;
 import org.apache.yetus.audience.InterfaceAudience;
@@ -321,20 +325,26 @@ protected void chooseSinks() {
       if (!useZk || ReplicationUtils.isPeerClusterSupportReplicationOffload(conn)) {
         useZk = false;
         slaveAddresses = fetchSlavesAddresses();
+        if (slaveAddresses.isEmpty()) {
+          LOG.warn("No sinks available at peer. Try fetch sinks by using zk.");
+          useZk = true;
+        }
       } else {
         useZk = true;
       }
     } catch (Throwable t) {
       LOG.warn("Peer {} try to fetch servers by admin failed. Using zk impl.", ctx.getPeerId(), t);
       useZk = true;
     }
+
     if (useZk) {
       slaveAddresses = fetchSlavesAddressesByZK();
     }
 
     if (slaveAddresses.isEmpty()) {
-      LOG.warn("No sinks available at peer. Will not be able to replicate");
+      LOG.warn("No sinks available at peer. Will not be able to replicate.");
     }
+
     Collections.shuffle(slaveAddresses, ThreadLocalRandom.current());
     int numSinks = (int) Math.ceil(slaveAddresses.size() * ratio);
     synchronized (this) {
@@ -368,10 +378,10 @@ protected SinkPeer getReplicationSink() throws IOException {
   }
 
   private SinkPeer createSinkPeer(ServerName serverName) throws IOException {
-    if (ReplicationUtils.isPeerClusterSupportReplicationOffload(conn)) {
-      return new ReplicationServerSinkPeer(serverName, conn.getReplicationServerAdmin(serverName));
-    } else {
+    if (fetchServersUseZk) {
       return new RegionServerSinkPeer(serverName, conn.getRegionServerAdmin(serverName));
+    } else {
+      return new ReplicationServerSinkPeer(serverName, conn.getReplicationServerAdmin(serverName));
     }
   }