Github user enixon commented on a diff in the pull request:
https://github.com/apache/zookeeper/pull/628#discussion_r231364616
--- Diff:
zookeeper-server/src/main/java/org/apache/zookeeper/server/quorum/ObserverMaster.java
---
@@ -0,0 +1,514 @@
+/**
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements. See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership. The ASF licenses this file
+ * to you 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.
+ */
+
+package org.apache.zookeeper.server.quorum;
+
+import org.apache.zookeeper.jmx.MBeanRegistry;
+import org.apache.zookeeper.server.Request;
+import org.apache.zookeeper.server.ZKDatabase;
+
+import java.io.BufferedInputStream;
+import java.io.ByteArrayInputStream;
+import java.io.DataInputStream;
+import java.io.IOException;
+import java.net.ServerSocket;
+import java.net.Socket;
+import java.net.SocketAddress;
+import java.util.ArrayList;
+import java.util.Arrays;
+import java.util.Collections;
+import java.util.Iterator;
+import java.util.Set;
+import java.util.concurrent.ConcurrentHashMap;
+import java.util.concurrent.ConcurrentLinkedQueue;
+import java.util.concurrent.Executors;
+import java.util.concurrent.ScheduledExecutorService;
+import java.util.concurrent.TimeUnit;
+import java.util.concurrent.atomic.AtomicLong;
+
+import org.apache.zookeeper.server.quorum.auth.QuorumAuthServer;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+/**
+ * Used by Followers to host Observers. This reduces the network load on
the Leader process by pushing
+ * the responsibility for keeping Observers in sync off the leading peer.
+ *
+ * It is expected that Observers will continue to perform the initial
vetting of clients and requests.
+ * Observers send the request to the follower where it is received by an
ObserverMaster.
+ *
+ * The ObserverMaster forwards a copy of the request to the ensemble
Leader and inserts it into its own
+ * request processor pipeline where it can be matched with the response
comes back. All commits received
+ * from the Leader will be forwarded along to every Learner connected to
the ObserverMaster.
+ *
+ * New Learners connecting to a Follower will receive a LearnerHandler
object and be party to its syncing logic
+ * to be brought up to date.
+ *
+ * The logic is quite a bit simpler than the corresponding logic in Leader
because it only hosts observers.
+ */
+public class ObserverMaster implements LearnerMaster, Runnable {
+ private static final Logger LOG =
LoggerFactory.getLogger(ObserverMaster.class);
+
+ //Follower counter
+ private final AtomicLong followerCounter = new AtomicLong(-1);
--- End diff --
This copies the followerCounter of the Leader as used in the beginning of
LearnerHandler::run. My understanding is that this provides backwards
compatibility for when the first FOLLOWERINFO or OBSERVERINFO packet from the
Learner does not supply a LearnerInfo and the Leader (or ObserverMaster here)
needs to come up with a unique server id for the new Learner with which to
refer to it in all internal collections.
It will also end up being used in #337 if/when that patch is accepted.
I'm happy to add explanatory comments for these counters in the Leader and
ObserverMaster but am unsure of how much of the above should be included. What
do you think of a simple "assigns unique server id for internal tracking of
Learners that do not provide a unique id"?
---
