Google Git
Sign in
apache / activemq-website / 9bef98a777362a978ca0249a9a9db4423fb294c9 / . / src / components / artemis / documentation / 1.5.3 / clusters.html
blob: a8398e23905a69f8174873436f907b601730abaf [file] [log] [blame]
<!DOCTYPE HTML>
<html lang="en" >
<head>
<meta charset="UTF-8">
<meta http-equiv="X-UA-Compatible" content="IE=edge" />
<title>Clusters | ActiveMQ Artemis Documentation</title>
<meta content="text/html; charset=utf-8" http-equiv="Content-Type">
<meta name="description" content="">
<meta name="generator" content="GitBook 2.1.0">
<meta name="HandheldFriendly" content="true"/>
<meta name="viewport" content="width=device-width, initial-scale=1, user-scalable=no">
<meta name="apple-mobile-web-app-capable" content="yes">
<meta name="apple-mobile-web-app-status-bar-style" content="black">
<link rel="apple-touch-icon-precomposed" sizes="152x152" href="gitbook/images/apple-touch-icon-precomposed-152.png">
<link rel="shortcut icon" href="gitbook/images/favicon.ico" type="image/x-icon">
<link rel="stylesheet" href="gitbook/style.css">
<link rel="next" href="./ha.html" />
<link rel="prev" href="./duplicate-detection.html" />
</head>
<body>
<div class="book" data-level="39" data-basepath="." data-revision="Wed Sep 23 2015 12:51:51 GMT+0100 (BST)">
<div class="book-summary">
<div class="book-search">
<input type="text" placeholder="Type to search" class="form-control" />
</div>
<ul class="summary">
<li class="chapter " data-level="0" data-path="index.html">
<a href="./index.html">
<i class="fa fa-check"></i>
Introduction
</a>
</li>
<li class="chapter " data-level="1" data-path="notice.html">
<a href="./notice.html">
<i class="fa fa-check"></i>
<b>1.</b>
Legal Notice
</a>
</li>
<li class="chapter " data-level="2" data-path="preface.html">
<a href="./preface.html">
<i class="fa fa-check"></i>
<b>2.</b>
Preface
</a>
</li>
<li class="chapter " data-level="3" data-path="project-info.html">
<a href="./project-info.html">
<i class="fa fa-check"></i>
<b>3.</b>
Project Info
</a>
</li>
<li class="chapter " data-level="4" data-path="messaging-concepts.html">
<a href="./messaging-concepts.html">
<i class="fa fa-check"></i>
<b>4.</b>
Messaging Concepts
</a>
</li>
<li class="chapter " data-level="5" data-path="architecture.html">
<a href="./architecture.html">
<i class="fa fa-check"></i>
<b>5.</b>
Architecture
</a>
</li>
<li class="chapter " data-level="6" data-path="using-server.html">
<a href="./using-server.html">
<i class="fa fa-check"></i>
<b>6.</b>
Using the Server
</a>
</li>
<li class="chapter " data-level="7" data-path="using-jms.html">
<a href="./using-jms.html">
<i class="fa fa-check"></i>
<b>7.</b>
Using JMS
</a>
</li>
<li class="chapter " data-level="8" data-path="using-core.html">
<a href="./using-core.html">
<i class="fa fa-check"></i>
<b>8.</b>
Using Core
</a>
</li>
<li class="chapter " data-level="9" data-path="jms-core-mapping.html">
<a href="./jms-core-mapping.html">
<i class="fa fa-check"></i>
<b>9.</b>
Mapping JMS Concepts to the Core API
</a>
</li>
<li class="chapter " data-level="10" data-path="client-classpath.html">
<a href="./client-classpath.html">
<i class="fa fa-check"></i>
<b>10.</b>
The Client Classpath
</a>
</li>
<li class="chapter " data-level="11" data-path="examples.html">
<a href="./examples.html">
<i class="fa fa-check"></i>
<b>11.</b>
Examples
</a>
</li>
<li class="chapter " data-level="12" data-path="wildcard-routing.html">
<a href="./wildcard-routing.html">
<i class="fa fa-check"></i>
<b>12.</b>
Routing Messages With Wild Cards
</a>
</li>
<li class="chapter " data-level="13" data-path="wildcard-syntax.html">
<a href="./wildcard-syntax.html">
<i class="fa fa-check"></i>
<b>13.</b>
Understanding the Apache ActiveMQ Artemis Wildcard Syntax
</a>
</li>
<li class="chapter " data-level="14" data-path="filter-expressions.html">
<a href="./filter-expressions.html">
<i class="fa fa-check"></i>
<b>14.</b>
Filter Expressions
</a>
</li>
<li class="chapter " data-level="15" data-path="persistence.html">
<a href="./persistence.html">
<i class="fa fa-check"></i>
<b>15.</b>
Persistence
</a>
</li>
<li class="chapter " data-level="16" data-path="configuring-transports.html">
<a href="./configuring-transports.html">
<i class="fa fa-check"></i>
<b>16.</b>
Configuring Transports
</a>
</li>
<li class="chapter " data-level="17" data-path="connection-ttl.html">
<a href="./connection-ttl.html">
<i class="fa fa-check"></i>
<b>17.</b>
Detecting Dead Connections
</a>
</li>
<li class="chapter " data-level="18" data-path="slow-consumers.html">
<a href="./slow-consumers.html">
<i class="fa fa-check"></i>
<b>18.</b>
Detecting Slow Consumers
</a>
</li>
<li class="chapter " data-level="19" data-path="transaction-config.html">
<a href="./transaction-config.html">
<i class="fa fa-check"></i>
<b>19.</b>
Resource Manager Configuration
</a>
</li>
<li class="chapter " data-level="20" data-path="flow-control.html">
<a href="./flow-control.html">
<i class="fa fa-check"></i>
<b>20.</b>
Flow Control
</a>
</li>
<li class="chapter " data-level="21" data-path="send-guarantees.html">
<a href="./send-guarantees.html">
<i class="fa fa-check"></i>
<b>21.</b>
Guarantees of sends and commits
</a>
</li>
<li class="chapter " data-level="22" data-path="undelivered-messages.html">
<a href="./undelivered-messages.html">
<i class="fa fa-check"></i>
<b>22.</b>
Message Redelivery and Undelivered Messages
</a>
</li>
<li class="chapter " data-level="23" data-path="message-expiry.html">
<a href="./message-expiry.html">
<i class="fa fa-check"></i>
<b>23.</b>
Message Expiry
</a>
</li>
<li class="chapter " data-level="24" data-path="large-messages.html">
<a href="./large-messages.html">
<i class="fa fa-check"></i>
<b>24.</b>
Large Messages
</a>
</li>
<li class="chapter " data-level="25" data-path="paging.html">
<a href="./paging.html">
<i class="fa fa-check"></i>
<b>25.</b>
Paging
</a>
</li>
<li class="chapter " data-level="26" data-path="queue-attributes.html">
<a href="./queue-attributes.html">
<i class="fa fa-check"></i>
<b>26.</b>
Queue Attributes
</a>
</li>
<li class="chapter " data-level="27" data-path="scheduled-messages.html">
<a href="./scheduled-messages.html">
<i class="fa fa-check"></i>
<b>27.</b>
Scheduled Messages
</a>
</li>
<li class="chapter " data-level="28" data-path="last-value-queues.html">
<a href="./last-value-queues.html">
<i class="fa fa-check"></i>
<b>28.</b>
Last-Value Queues
</a>
</li>
<li class="chapter " data-level="29" data-path="message-grouping.html">
<a href="./message-grouping.html">
<i class="fa fa-check"></i>
<b>29.</b>
Message Grouping
</a>
</li>
<li class="chapter " data-level="30" data-path="pre-acknowledge.html">
<a href="./pre-acknowledge.html">
<i class="fa fa-check"></i>
<b>30.</b>
Extra Acknowledge Modes
</a>
</li>
<li class="chapter " data-level="31" data-path="management.html">
<a href="./management.html">
<i class="fa fa-check"></i>
<b>31.</b>
Management
</a>
</li>
<li class="chapter " data-level="32" data-path="security.html">
<a href="./security.html">
<i class="fa fa-check"></i>
<b>32.</b>
Security
</a>
</li>
<li class="chapter " data-level="33" data-path="resource-limits.html">
<a href="./resource-limits.html">
<i class="fa fa-check"></i>
<b>33.</b>
Resource Limits
</a>
</li>
<li class="chapter " data-level="34" data-path="jms-bridge.html">
<a href="./jms-bridge.html">
<i class="fa fa-check"></i>
<b>34.</b>
The JMS Bridge
</a>
</li>
<li class="chapter " data-level="35" data-path="client-reconnection.html">
<a href="./client-reconnection.html">
<i class="fa fa-check"></i>
<b>35.</b>
Client Reconnection and Session Reattachment
</a>
</li>
<li class="chapter " data-level="36" data-path="diverts.html">
<a href="./diverts.html">
<i class="fa fa-check"></i>
<b>36.</b>
Diverting and Splitting Message Flows
</a>
</li>
<li class="chapter " data-level="37" data-path="core-bridges.html">
<a href="./core-bridges.html">
<i class="fa fa-check"></i>
<b>37.</b>
Core Bridges
</a>
</li>
<li class="chapter " data-level="38" data-path="duplicate-detection.html">
<a href="./duplicate-detection.html">
<i class="fa fa-check"></i>
<b>38.</b>
Duplicate Message Detection
</a>
</li>
<li class="chapter active" data-level="39" data-path="clusters.html">
<a href="./clusters.html">
<i class="fa fa-check"></i>
<b>39.</b>
Clusters
</a>
</li>
<li class="chapter " data-level="40" data-path="ha.html">
<a href="./ha.html">
<i class="fa fa-check"></i>
<b>40.</b>
High Availability and Failover
</a>
</li>
<li class="chapter " data-level="41" data-path="graceful-shutdown.html">
<a href="./graceful-shutdown.html">
<i class="fa fa-check"></i>
<b>41.</b>
Graceful Server Shutdown
</a>
</li>
<li class="chapter " data-level="42" data-path="libaio.html">
<a href="./libaio.html">
<i class="fa fa-check"></i>
<b>42.</b>
Libaio Native Libraries
</a>
</li>
<li class="chapter " data-level="43" data-path="thread-pooling.html">
<a href="./thread-pooling.html">
<i class="fa fa-check"></i>
<b>43.</b>
Thread management
</a>
</li>
<li class="chapter " data-level="44" data-path="logging.html">
<a href="./logging.html">
<i class="fa fa-check"></i>
<b>44.</b>
Logging
</a>
</li>
<li class="chapter " data-level="45" data-path="rest.html">
<a href="./rest.html">
<i class="fa fa-check"></i>
<b>45.</b>
REST Interface
</a>
</li>
<li class="chapter " data-level="46" data-path="embedding-activemq.html">
<a href="./embedding-activemq.html">
<i class="fa fa-check"></i>
<b>46.</b>
Embedding Apache ActiveMQ Artemis
</a>
</li>
<li class="chapter " data-level="47" data-path="spring-integration.html">
<a href="./spring-integration.html">
<i class="fa fa-check"></i>
<b>47.</b>
Spring Integration
</a>
</li>
<li class="chapter " data-level="48" data-path="aerogear-integration.html">
<a href="./aerogear-integration.html">
<i class="fa fa-check"></i>
<b>48.</b>
AeroGear Integration
</a>
</li>
<li class="chapter " data-level="49" data-path="vertx-integration.html">
<a href="./vertx-integration.html">
<i class="fa fa-check"></i>
<b>49.</b>
VertX Integration
</a>
</li>
<li class="chapter " data-level="50" data-path="intercepting-operations.html">
<a href="./intercepting-operations.html">
<i class="fa fa-check"></i>
<b>50.</b>
Intercepting Operations
</a>
</li>
<li class="chapter " data-level="51" data-path="protocols-interoperability.html">
<a href="./protocols-interoperability.html">
<i class="fa fa-check"></i>
<b>51.</b>
Protocols and Interoperability
</a>
</li>
<li class="chapter " data-level="52" data-path="tools.html">
<a href="./tools.html">
<i class="fa fa-check"></i>
<b>52.</b>
Tools
</a>
</li>
<li class="chapter " data-level="53" data-path="maven-plugin.html">
<a href="./maven-plugin.html">
<i class="fa fa-check"></i>
<b>53.</b>
Maven Plugin
</a>
</li>
<li class="chapter " data-level="54" data-path="perf-tuning.html">
<a href="./perf-tuning.html">
<i class="fa fa-check"></i>
<b>54.</b>
Troubleshooting and Performance Tuning
</a>
</li>
<li class="chapter " data-level="55" data-path="configuration-index.html">
<a href="./configuration-index.html">
<i class="fa fa-check"></i>
<b>55.</b>
Configuration Reference
</a>
</li>
<li class="divider"></li>
<li>
<a href="https://www.gitbook.com" target="blank" class="gitbook-link">
Published with GitBook
</a>
</li>
</ul>
</div>
<div class="book-body">
<div class="body-inner">
<div class="book-header">
<!-- Actions Left -->
<a href="#" class="btn pull-left toggle-summary" aria-label="Table of Contents"><i class="fa fa-align-justify"></i></a>
<a href="#" class="btn pull-left toggle-search" aria-label="Search"><i class="fa fa-search"></i></a>
<div id="font-settings-wrapper" class="dropdown pull-left">
<a href="#" class="btn toggle-dropdown" aria-label="Font Settings"><i class="fa fa-font"></i>
</a>
<div class="dropdown-menu font-settings">
<div class="dropdown-caret">
<span class="caret-outer"></span>
<span class="caret-inner"></span>
</div>
<div class="buttons">
<button type="button" id="reduce-font-size" class="button size-2">A</button>
<button type="button" id="enlarge-font-size" class="button size-2">A</button>
</div>
<div class="buttons font-family-list">
<button type="button" data-font="0" class="button">Serif</button>
<button type="button" data-font="1" class="button">Sans</button>
</div>
<div class="buttons color-theme-list">
<button type="button" id="color-theme-preview-0" class="button size-3" data-theme="0">White</button>
<button type="button" id="color-theme-preview-1" class="button size-3" data-theme="1">Sepia</button>
<button type="button" id="color-theme-preview-2" class="button size-3" data-theme="2">Night</button>
</div>
</div>
</div>
<!-- Actions Right -->
<div class="dropdown pull-right">
<a href="#" class="btn toggle-dropdown" aria-label="Share"><i class="fa fa-share-alt"></i>
</a>
<div class="dropdown-menu font-settings dropdown-left">
<div class="dropdown-caret">
<span class="caret-outer"></span>
<span class="caret-inner"></span>
</div>
<div class="buttons">
<button type="button" data-sharing="twitter" class="button">
Share on Twitter
</button>
<button type="button" data-sharing="google-plus" class="button">
Share on Google
</button>
<button type="button" data-sharing="facebook" class="button">
Share on Facebook
</button>
<button type="button" data-sharing="weibo" class="button">
Share on Weibo
</button>
<button type="button" data-sharing="instapaper" class="button">
Share on Instapaper
</button>
</div>
</div>
</div>
<a href="#" target="_blank" class="btn pull-right google-plus-sharing-link sharing-link" data-sharing="google-plus" aria-label="Google"><i class="fa fa-google-plus"></i></a>
<a href="#" target="_blank" class="btn pull-right facebook-sharing-link sharing-link" data-sharing="facebook" aria-label="Facebook"><i class="fa fa-facebook"></i></a>
<a href="#" target="_blank" class="btn pull-right twitter-sharing-link sharing-link" data-sharing="twitter" aria-label="Twitter"><i class="fa fa-twitter"></i></a>
<!-- Title -->
<h1>
<i class="fa fa-circle-o-notch fa-spin"></i>
<a href="./" >ActiveMQ Artemis Documentation</a>
</h1>
</div>
<div class="page-wrapper" tabindex="-1">
<div class="page-inner">
<section class="normal" id="section-">
<h1 id="clusters">Clusters</h1>
<h2 id="clusters-overview">Clusters Overview</h2>
<p>Apache ActiveMQ Artemis clusters allow groups of Apache ActiveMQ Artemis servers to be grouped
together in order to share message processing load. Each active node in
the cluster is an active Apache ActiveMQ Artemis server which manages its own messages
and handles its own connections.</p>
<p>The cluster is formed by each node declaring <em>cluster connections</em> to
other nodes in the core configuration file <code>broker.xml</code>.
When a node forms a cluster connection to another node, internally it
creates a <em>core bridge</em> (as described in <a href="core-bridges.html">Core Bridges</a>) connection between it and
the other node, this is done transparently behind the scenes - you don&apos;t
have to declare an explicit bridge for each node. These cluster
connections allow messages to flow between the nodes of the cluster to
balance load.</p>
<p>Nodes can be connected together to form a cluster in many different
topologies, we will discuss a couple of the more common topologies later
in this chapter.</p>
<p>We&apos;ll also discuss client side load balancing, where we can balance
client connections across the nodes of the cluster, and we&apos;ll consider
message redistribution where Apache ActiveMQ Artemis will redistribute messages between
nodes to avoid starvation.</p>
<p>Another important part of clustering is <em>server discovery</em> where servers
can broadcast their connection details so clients or other servers can
connect to them with the minimum of configuration.</p>
<blockquote>
<p><strong>Warning</strong></p>
<p>Once a cluster node has been configured it is common to simply copy
that configuration to other nodes to produce a symmetric cluster.
However, care must be taken when copying the Apache ActiveMQ Artemis files. Do not
copy the Apache ActiveMQ Artemis <em>data</em> (i.e. the <code>bindings</code>, <code>journal</code>, and
<code>large-messages</code> directories) from one node to another. When a node is
started for the first time and initializes its journal files it also
persists a special identifier to the <code>journal</code> directory. This id
<em>must</em> be unique among nodes in the cluster or the cluster will not
form properly.</p>
</blockquote>
<h2 id="server-discovery">Server discovery</h2>
<p>Server discovery is a mechanism by which servers can propagate their
connection details to:</p>
<ul>
<li><p>Messaging clients. A messaging client wants to be able to connect to
the servers of the cluster without having specific knowledge of
which servers in the cluster are up at any one time.</p>
</li>
<li><p>Other servers. Servers in a cluster want to be able to create
cluster connections to each other without having prior knowledge of
all the other servers in the cluster.</p>
</li>
</ul>
<p>This information, let&apos;s call it the Cluster Topology, is actually sent
around normal Apache ActiveMQ Artemis connections to clients and to other servers over
cluster connections. This being the case we need a way of establishing
the initial first connection. This can be done using dynamic discovery
techniques like
<a href="http://en.wikipedia.org/wiki/User_Datagram_Protocol" target="_blank">UDP</a> and
<a href="http://www.jgroups.org/" target="_blank">JGroups</a>, or by providing a list of initial
connectors.</p>
<h3 id="dynamic-discovery">Dynamic Discovery</h3>
<p>Server discovery uses
<a href="http://en.wikipedia.org/wiki/User_Datagram_Protocol" target="_blank">UDP</a> multicast or
<a href="http://www.jgroups.org/" target="_blank">JGroups</a> to broadcast server connection
settings.</p>
<h4 id="broadcast-groups">Broadcast Groups</h4>
<p>A broadcast group is the means by which a server broadcasts connectors
over the network. A connector defines a way in which a client (or other
server) can make connections to the server. For more information on what
a connector is, please see <a href="configuring-transports.html">Configuring the Transport</a>.</p>
<p>The broadcast group takes a set of connector pairs, each connector pair
contains connection settings for a live and backup server (if one
exists) and broadcasts them on the network. Depending on which
broadcasting technique you configure the cluster, it uses either UDP or
JGroups to broadcast connector pairs information.</p>
<p>Broadcast groups are defined in the server configuration file
<code>broker.xml</code>. There can be many broadcast groups per
Apache ActiveMQ Artemis server. All broadcast groups must be defined in a
<code>broadcast-groups</code> element.</p>
<p>Let&apos;s take a look at an example broadcast group from
<code>broker.xml</code> that defines a UDP broadcast group:</p>
<pre><code>&lt;broadcast-groups&gt;
&lt;broadcast-group name=&quot;my-broadcast-group&quot;&gt;
&lt;local-bind-address&gt;172.16.9.3&lt;/local-bind-address&gt;
&lt;local-bind-port&gt;5432&lt;/local-bind-port&gt;
&lt;group-address&gt;231.7.7.7&lt;/group-address&gt;
&lt;group-port&gt;9876&lt;/group-port&gt;
&lt;broadcast-period&gt;2000&lt;/broadcast-period&gt;
&lt;connector-ref connector-name=&quot;netty-connector&quot;/&gt;
&lt;/broadcast-group&gt;
&lt;/broadcast-groups&gt;
</code></pre><p>Some of the broadcast group parameters are optional and you&apos;ll normally
use the defaults, but we specify them all in the above example for
clarity. Let&apos;s discuss each one in turn:</p>
<ul>
<li><p><code>name</code> attribute. Each broadcast group in the server must have a
unique name.</p>
</li>
<li><p><code>local-bind-address</code>. This is the local bind address that the
datagram socket is bound to. If you have multiple network interfaces
on your server, you would specify which one you wish to use for
broadcasts by setting this property. If this property is not
specified then the socket will be bound to the wildcard address, an
IP address chosen by the kernel. This is a UDP specific attribute.</p>
</li>
<li><p><code>local-bind-port</code>. If you want to specify a local port to which the
datagram socket is bound you can specify it here. Normally you would
just use the default value of <code>-1</code> which signifies that an anonymous
port should be used. This parameter is always specified in
conjunction with <code>local-bind-address</code>. This is a UDP specific
attribute.</p>
</li>
<li><p><code>group-address</code>. This is the multicast address to which the data
will be broadcast. It is a class D IP address in the range
<code>224.0.0.0</code> to <code>239.255.255.255</code>, inclusive. The address <code>224.0.0.0</code>
is reserved and is not available for use. This parameter is
mandatory. This is a UDP specific attribute.</p>
</li>
<li><p><code>group-port</code>. This is the UDP port number used for broadcasting.
This parameter is mandatory. This is a UDP specific attribute.</p>
</li>
<li><p><code>broadcast-period</code>. This is the period in milliseconds between
consecutive broadcasts. This parameter is optional, the default
value is <code>2000</code> milliseconds.</p>
</li>
<li><p><code>connector-ref</code>. This specifies the connector and optional backup
connector that will be broadcasted (see <a href="configuring-transports.html">Configuring the Transport</a> for more information on
connectors). The connector to be broadcasted is specified by the
<code>connector-name</code> attribute.</p>
</li>
</ul>
<p>Here is another example broadcast group that defines a JGroups broadcast
group:</p>
<pre><code>&lt;broadcast-groups&gt;
&lt;broadcast-group name=&quot;my-broadcast-group&quot;&gt;
&lt;jgroups-file&gt;test-jgroups-file_ping.xml&lt;/jgroups-file&gt;
&lt;jgroups-channel&gt;activemq_broadcast_channel&lt;/jgroups-channel&gt;
&lt;broadcast-period&gt;2000&lt;/broadcast-period&gt;
&lt;connector-ref connector-name=&quot;netty-connector&quot;/&gt;
&lt;/broadcast-group&gt;
&lt;/broadcast-groups&gt;
</code></pre><p>To be able to use JGroups to broadcast, one must specify two attributes,
i.e. <code>jgroups-file</code> and <code>jgroups-channel</code>, as discussed in details as
following:</p>
<ul>
<li><p><code>jgroups-file</code> attribute. This is the name of JGroups configuration
file. It will be used to initialize JGroups channels. Make sure the
file is in the java resource path so that Apache ActiveMQ Artemis can load it.</p>
</li>
<li><p><code>jgroups-channel</code> attribute. The name that JGroups channels connect
to for broadcasting.</p>
</li>
</ul>
<blockquote>
<p><strong>Note</strong></p>
<p>The JGroups attributes (<code>jgroups-file</code> and <code>jgroups-channel</code>) and UDP
specific attributes described above are exclusive of each other. Only
one set can be specified in a broadcast group configuration. Don&apos;t mix
them!</p>
</blockquote>
<p>The following is an example of a JGroups file</p>
<pre><code>&lt;config xmlns=&quot;urn:org:jgroups&quot;
xmlns:xsi=&quot;http://www.w3.org/2001/XMLSchema-instance&quot;
xsi:schemaLocation=&quot;urn:org:jgroups http://www.jgroups.org/schema/JGroups-3.0.xsd&quot;&gt;
&lt;TCP loopback=&quot;true&quot;
recv_buf_size=&quot;20000000&quot;
send_buf_size=&quot;640000&quot;
discard_incompatible_packets=&quot;true&quot;
max_bundle_size=&quot;64000&quot;
max_bundle_timeout=&quot;30&quot;
enable_bundling=&quot;true&quot;
use_send_queues=&quot;false&quot;
sock_conn_timeout=&quot;300&quot;
thread_pool.enabled=&quot;true&quot;
thread_pool.min_threads=&quot;1&quot;
thread_pool.max_threads=&quot;10&quot;
thread_pool.keep_alive_time=&quot;5000&quot;
thread_pool.queue_enabled=&quot;false&quot;
thread_pool.queue_max_size=&quot;100&quot;
thread_pool.rejection_policy=&quot;run&quot;
oob_thread_pool.enabled=&quot;true&quot;
oob_thread_pool.min_threads=&quot;1&quot;
oob_thread_pool.max_threads=&quot;8&quot;
oob_thread_pool.keep_alive_time=&quot;5000&quot;
oob_thread_pool.queue_enabled=&quot;false&quot;
oob_thread_pool.queue_max_size=&quot;100&quot;
oob_thread_pool.rejection_policy=&quot;run&quot;/&gt;
&lt;FILE_PING location=&quot;../file.ping.dir&quot;/&gt;
&lt;MERGE2 max_interval=&quot;30000&quot;
min_interval=&quot;10000&quot;/&gt;
&lt;FD_SOCK/&gt;
&lt;FD timeout=&quot;10000&quot; max_tries=&quot;5&quot; /&gt;
&lt;VERIFY_SUSPECT timeout=&quot;1500&quot; /&gt;
&lt;BARRIER /&gt;
&lt;pbcast.NAKACK
use_mcast_xmit=&quot;false&quot;
retransmit_timeout=&quot;300,600,1200,2400,4800&quot;
discard_delivered_msgs=&quot;true&quot;/&gt;
&lt;UNICAST timeout=&quot;300,600,1200&quot; /&gt;
&lt;pbcast.STABLE stability_delay=&quot;1000&quot; desired_avg_gossip=&quot;50000&quot;
max_bytes=&quot;400000&quot;/&gt;
&lt;pbcast.GMS print_local_addr=&quot;true&quot; join_timeout=&quot;3000&quot;
view_bundling=&quot;true&quot;/&gt;
&lt;FC max_credits=&quot;2000000&quot;
min_threshold=&quot;0.10&quot;/&gt;
&lt;FRAG2 frag_size=&quot;60000&quot; /&gt;
&lt;pbcast.STATE_TRANSFER/&gt;
&lt;pbcast.FLUSH timeout=&quot;0&quot;/&gt;
&lt;/config&gt;
</code></pre><p>As it shows, the file content defines a jgroups protocol stacks. If you
want Apache ActiveMQ Artemis to use this stacks for channel creation, you have to make
sure the value of <code>jgroups-file</code> in your broadcast-group/discovery-group
configuration to be the name of this jgroups configuration file. For
example if the above stacks configuration is stored in a file named
&quot;jgroups-stacks.xml&quot; then your <code>jgroups-file</code> should be like</p>
<pre><code>&lt;jgroups-file&gt;jgroups-stacks.xml&lt;/jgroups-file&gt;
</code></pre><h4 id="discovery-groups">Discovery Groups</h4>
<p>While the broadcast group defines how connector information is
broadcasted from a server, a discovery group defines how connector
information is received from a broadcast endpoint (a UDP multicast
address or JGroup channel).</p>
<p>A discovery group maintains a list of connector pairs - one for each
broadcast by a different server. As it receives broadcasts on the
broadcast endpoint from a particular server it updates its entry in the
list for that server.</p>
<p>If it has not received a broadcast from a particular server for a length
of time it will remove that server&apos;s entry from its list.</p>
<p>Discovery groups are used in two places in Apache ActiveMQ Artemis:</p>
<ul>
<li><p>By cluster connections so they know how to obtain an initial
connection to download the topology</p>
</li>
<li><p>By messaging clients so they know how to obtain an initial
connection to download the topology</p>
</li>
</ul>
<p>Although a discovery group will always accept broadcasts, its current
list of available live and backup servers is only ever used when an
initial connection is made, from then server discovery is done over the
normal Apache ActiveMQ Artemis connections.</p>
<blockquote>
<p><strong>Note</strong></p>
<p>Each discovery group must be configured with broadcast endpoint (UDP
or JGroups) that matches its broadcast group counterpart. For example,
if broadcast is configured using UDP, the discovery group must also
use UDP, and the same multicast address.</p>
</blockquote>
<h4 id="defining-discovery-groups-on-the-server">Defining Discovery Groups on the Server</h4>
<p>For cluster connections, discovery groups are defined in the server side
configuration file <code>broker.xml</code>. All discovery groups
must be defined inside a <code>discovery-groups</code> element. There can be many
discovery groups defined by Apache ActiveMQ Artemis server. Let&apos;s look at an example:</p>
<pre><code>&lt;discovery-groups&gt;
&lt;discovery-group name=&quot;my-discovery-group&quot;&gt;
&lt;local-bind-address&gt;172.16.9.7&lt;/local-bind-address&gt;
&lt;group-address&gt;231.7.7.7&lt;/group-address&gt;
&lt;group-port&gt;9876&lt;/group-port&gt;
&lt;refresh-timeout&gt;10000&lt;/refresh-timeout&gt;
&lt;/discovery-group&gt;
&lt;/discovery-groups&gt;
</code></pre><p>We&apos;ll consider each parameter of the discovery group:</p>
<ul>
<li><p><code>name</code> attribute. Each discovery group must have a unique name per
server.</p>
</li>
<li><p><code>local-bind-address</code>. If you are running with multiple network
interfaces on the same machine, you may want to specify that the
discovery group listens only only a specific interface. To do this
you can specify the interface address with this parameter. This
parameter is optional. This is a UDP specific attribute.</p>
</li>
<li><p><code>group-address</code>. This is the multicast IP address of the group to
listen on. It should match the <code>group-address</code> in the broadcast
group that you wish to listen from. This parameter is mandatory.
This is a UDP specific attribute.</p>
</li>
<li><p><code>group-port</code>. This is the UDP port of the multicast group. It should
match the <code>group-port</code> in the broadcast group that you wish to
listen from. This parameter is mandatory. This is a UDP specific
attribute.</p>
</li>
<li><p><code>refresh-timeout</code>. This is the period the discovery group waits
after receiving the last broadcast from a particular server before
removing that servers connector pair entry from its list. You would
normally set this to a value significantly higher than the
<code>broadcast-period</code> on the broadcast group otherwise servers might
intermittently disappear from the list even though they are still
broadcasting due to slight differences in timing. This parameter is
optional, the default value is <code>10000</code> milliseconds (10 seconds).</p>
</li>
</ul>
<p>Here is another example that defines a JGroups discovery group:</p>
<pre><code>&lt;discovery-groups&gt;
&lt;discovery-group name=&quot;my-broadcast-group&quot;&gt;
&lt;jgroups-file&gt;test-jgroups-file_ping.xml&lt;/jgroups-file&gt;
&lt;jgroups-channel&gt;activemq_broadcast_channel&lt;/jgroups-channel&gt;
&lt;refresh-timeout&gt;10000&lt;/refresh-timeout&gt;
&lt;/discovery-group&gt;
&lt;/discovery-groups&gt;
</code></pre><p>To receive broadcast from JGroups channels, one must specify two
attributes, <code>jgroups-file</code> and <code>jgroups-channel</code>, as discussed in
details as following:</p>
<ul>
<li><p><code>jgroups-file</code> attribute. This is the name of JGroups configuration
file. It will be used to initialize JGroups channels. Make sure the
file is in the java resource path so that Apache ActiveMQ Artemis can load it.</p>
</li>
<li><p><code>jgroups-channel</code> attribute. The name that JGroups channels connect
to for receiving broadcasts.</p>
</li>
</ul>
<blockquote>
<p><strong>Note</strong></p>
<p>The JGroups attributes (<code>jgroups-file</code> and <code>jgroups-channel</code>) and UDP
specific attributes described above are exclusive of each other. Only
one set can be specified in a discovery group configuration. Don&apos;t mix
them!</p>
</blockquote>
<h4 id="discovery-groups-on-the-client-side">Discovery Groups on the Client Side</h4>
<p>Let&apos;s discuss how to configure an Apache ActiveMQ Artemis client to use discovery to
discover a list of servers to which it can connect. The way to do this
differs depending on whether you&apos;re using JMS or the core API.</p>
<h5 id="configuring-client-discovery-using-jms">Configuring client discovery using JMS</h5>
<p>If you&apos;re using JMS and you&apos;re using JNDI on the client to look up your
JMS connection factory instances then you can specify these parameters
in the JNDI context environment. e.g. in <code>jndi.properties</code>. Simply
ensure the host:port combination matches the group-address and
group-port from the corresponding <code>broadcast-group</code> on the server. Let&apos;s
take a look at an example:</p>
<pre><code>java.naming.factory.initial = ActiveMQInitialContextFactory
connectionFactory.myConnectionFactory=udp://231.7.7.7:9876
</code></pre><p>The element <code>discovery-group-ref</code> specifies the name of a discovery
group defined in <code>broker.xml</code>.</p>
<p>When this connection factory is downloaded from JNDI by a client
application and JMS connections are created from it, those connections
will be load-balanced across the list of servers that the discovery
group maintains by listening on the multicast address specified in the
discovery group configuration.</p>
<p>If you&apos;re using JMS, but you&apos;re not using JNDI to lookup a connection
factory - you&apos;re instantiating the JMS connection factory directly then
you can specify the discovery group parameters directly when creating
the JMS connection factory. Here&apos;s an example:</p>
<pre><code class="lang-java"><span class="hljs-keyword">final</span> String groupAddress = <span class="hljs-string">&quot;231.7.7.7&quot;</span>;
<span class="hljs-keyword">final</span> <span class="hljs-keyword">int</span> groupPort = <span class="hljs-number">9876</span>;
ConnectionFactory jmsConnectionFactory =
ActiveMQJMSClient.createConnectionFactory(<span class="hljs-keyword">new</span> DiscoveryGroupConfiguration(groupAddress, groupPort,
<span class="hljs-keyword">new</span> UDPBroadcastGroupConfiguration(groupAddress, groupPort, <span class="hljs-keyword">null</span>, -<span class="hljs-number">1</span>)), JMSFactoryType.CF);
Connection jmsConnection1 = jmsConnectionFactory.createConnection();
Connection jmsConnection2 = jmsConnectionFactory.createConnection();
</code></pre>
<p>The <code>refresh-timeout</code> can be set directly on the
DiscoveryGroupConfiguration by using the setter method
<code>setDiscoveryRefreshTimeout()</code> if you want to change the default value.</p>
<p>There is also a further parameter settable on the
DiscoveryGroupConfiguration using the setter method
<code>setDiscoveryInitialWaitTimeout()</code>. If the connection factory is used
immediately after creation then it may not have had enough time to
received broadcasts from all the nodes in the cluster. On first usage,
the connection factory will make sure it waits this long since creation
before creating the first connection. The default value for this
parameter is <code>10000</code> milliseconds.</p>
<h5 id="configuring-client-discovery-using-core">Configuring client discovery using Core</h5>
<p>If you&apos;re using the core API to directly instantiate
<code>ClientSessionFactory</code> instances, then you can specify the discovery
group parameters directly when creating the session factory. Here&apos;s an
example:</p>
<pre><code class="lang-java"><span class="hljs-keyword">final</span> String groupAddress = <span class="hljs-string">&quot;231.7.7.7&quot;</span>;
<span class="hljs-keyword">final</span> <span class="hljs-keyword">int</span> groupPort = <span class="hljs-number">9876</span>;
ServerLocator factory = ActiveMQClient.createServerLocatorWithHA(<span class="hljs-keyword">new</span> DiscoveryGroupConfiguration(groupAddress, groupPort,
<span class="hljs-keyword">new</span> UDPBroadcastGroupConfiguration(groupAddress, groupPort, <span class="hljs-keyword">null</span>, -<span class="hljs-number">1</span>))));
ClientSessionFactory factory = locator.createSessionFactory();
ClientSession session1 = factory.createSession();
ClientSession session2 = factory.createSession();
</code></pre>
<p>The <code>refresh-timeout</code> can be set directly on the
DiscoveryGroupConfiguration by using the setter method
<code>setDiscoveryRefreshTimeout()</code> if you want to change the default value.</p>
<p>There is also a further parameter settable on the
DiscoveryGroupConfiguration using the setter method
<code>setDiscoveryInitialWaitTimeout()</code>. If the session factory is used
immediately after creation then it may not have had enough time to
received broadcasts from all the nodes in the cluster. On first usage,
the session factory will make sure it waits this long since creation
before creating the first session. The default value for this parameter
is <code>10000</code> milliseconds.</p>
<h3 id="discovery-using-static-connectors">Discovery using static Connectors</h3>
<p>Sometimes it may be impossible to use UDP on the network you are using.
In this case its possible to configure a connection with an initial list
if possible servers. This could be just one server that you know will
always be available or a list of servers where at least one will be
available.</p>
<p>This doesn&apos;t mean that you have to know where all your servers are going
to be hosted, you can configure these servers to use the reliable
servers to connect to. Once they are connected there connection details
will be propagated via the server it connects to</p>
<h4 id="configuring-a-cluster-connection">Configuring a Cluster Connection</h4>
<p>For cluster connections there is no extra configuration needed, you just
need to make sure that any connectors are defined in the usual manner,
(see <a href="configuring-transports.html">Configuring the Transport</a> for more information on connectors). These are then referenced by
the cluster connection configuration.</p>
<h4 id="configuring-a-client-connection">Configuring a Client Connection</h4>
<p>A static list of possible servers can also be used by a normal client.</p>
<h5 id="configuring-client-discovery-using-jms">Configuring client discovery using JMS</h5>
<p>If you&apos;re using JMS and you&apos;re using JNDI on the client to look up your
JMS connection factory instances then you can specify these parameters
in the JNDI context environment in, e.g. <code>jndi.properties</code>:</p>
<pre><code>java.naming.factory.initial=org.apache.activemq.artemis.jndi.ActiveMQInitialContextFactory
connectionFactory.myConnectionFactory=(tcp://myhost:61616,tcp://myhost2:61616)
</code></pre><p>The <code>connectionFactory.myConnectionFactory</code> contains a list of servers to use for the
connection factory. When this connection factory used client application
and JMS connections are created from it, those connections will be
load-balanced across the list of servers defined within the brackets <code>()</code>.
The brackets are expanded so the same query cab be appended after the last bracket for ease.</p>
<p>If you&apos;re using JMS, but you&apos;re not using JNDI to lookup a connection
factory - you&apos;re instantiating the JMS connection factory directly then
you can specify the connector list directly when creating the JMS
connection factory. Here&apos;s an example:</p>
<pre><code class="lang-java">HashMap&lt;String, Object&gt; map = <span class="hljs-keyword">new</span> HashMap&lt;String, Object&gt;();
map.put(<span class="hljs-string">&quot;host&quot;</span>, <span class="hljs-string">&quot;myhost&quot;</span>);
map.put(<span class="hljs-string">&quot;port&quot;</span>, <span class="hljs-string">&quot;61616&quot;</span>);
TransportConfiguration server1 = <span class="hljs-keyword">new</span> TransportConfiguration(NettyConnectorFactory.class.getName(), map);
HashMap&lt;String, Object&gt; map2 = <span class="hljs-keyword">new</span> HashMap&lt;String, Object&gt;();
map2.put(<span class="hljs-string">&quot;host&quot;</span>, <span class="hljs-string">&quot;myhost2&quot;</span>);
map2.put(<span class="hljs-string">&quot;port&quot;</span>, <span class="hljs-string">&quot;61617&quot;</span>);
TransportConfiguration server2 = <span class="hljs-keyword">new</span> TransportConfiguration(NettyConnectorFactory.class.getName(), map2);
ActiveMQConnectionFactory cf = ActiveMQJMSClient.createConnectionFactoryWithHA(JMSFactoryType.CF, server1, server2);
</code></pre>
<h5 id="configuring-client-discovery-using-core">Configuring client discovery using Core</h5>
<p>If you are using the core API then the same can be done as follows:</p>
<pre><code class="lang-java">HashMap&lt;String, Object&gt; map = <span class="hljs-keyword">new</span> HashMap&lt;String, Object&gt;();
map.put(<span class="hljs-string">&quot;host&quot;</span>, <span class="hljs-string">&quot;myhost&quot;</span>);
map.put(<span class="hljs-string">&quot;port&quot;</span>, <span class="hljs-string">&quot;61616&quot;</span>);
TransportConfiguration server1 = <span class="hljs-keyword">new</span> TransportConfiguration(NettyConnectorFactory.class.getName(), map);
HashMap&lt;String, Object&gt; map2 = <span class="hljs-keyword">new</span> HashMap&lt;String, Object&gt;();
map2.put(<span class="hljs-string">&quot;host&quot;</span>, <span class="hljs-string">&quot;myhost2&quot;</span>);
map2.put(<span class="hljs-string">&quot;port&quot;</span>, <span class="hljs-string">&quot;61617&quot;</span>);
TransportConfiguration server2 = <span class="hljs-keyword">new</span> TransportConfiguration(NettyConnectorFactory.class.getName(), map2);
ServerLocator locator = ActiveMQClient.createServerLocatorWithHA(server1, server2);
ClientSessionFactory factory = locator.createSessionFactory();
ClientSession session = factory.createSession();
</code></pre>
<h2 id="server-side-message-load-balancing">Server-Side Message Load Balancing</h2>
<p>If cluster connections are defined between nodes of a cluster, then
Apache ActiveMQ Artemis will load balance messages arriving at a particular node from a
client.</p>
<p>Let&apos;s take a simple example of a cluster of four nodes A, B, C, and D
arranged in a <em>symmetric cluster</em> (described in Symmetrical Clusters section). We have a queue
called <code>OrderQueue</code> deployed on each node of the cluster.</p>
<p>We have client Ca connected to node A, sending orders to the server. We
have also have order processor clients Pa, Pb, Pc, and Pd connected to
each of the nodes A, B, C, D. If no cluster connection was defined on
node A, then as order messages arrive on node A they will all end up in
the <code>OrderQueue</code> on node A, so will only get consumed by the order
processor client attached to node A, Pa.</p>
<p>If we define a cluster connection on node A, then as ordered messages
arrive on node A instead of all of them going into the local
<code>OrderQueue</code> instance, they are distributed in a round-robin fashion
between all the nodes of the cluster. The messages are forwarded from
the receiving node to other nodes of the cluster. This is all done on
the server side, the client maintains a single connection to node A.</p>
<p>For example, messages arriving on node A might be distributed in the
following order between the nodes: B, D, C, A, B, D, C, A, B, D. The
exact order depends on the order the nodes started up, but the algorithm
used is round robin.</p>
<p>Apache ActiveMQ Artemis cluster connections can be configured to always blindly load
balance messages in a round robin fashion irrespective of whether there
are any matching consumers on other nodes, but they can be a bit
cleverer than that and also be configured to only distribute to other
nodes if they have matching consumers. We&apos;ll look at both these cases in
turn with some examples, but first we&apos;ll discuss configuring cluster
connections in general.</p>
<h3 id="configuring-cluster-connections">Configuring Cluster Connections</h3>
<p>Cluster connections group servers into clusters so that messages can be
load balanced between the nodes of the cluster. Let&apos;s take a look at a
typical cluster connection. Cluster connections are always defined in
<code>broker.xml</code> inside a <code>cluster-connection</code> element.
There can be zero or more cluster connections defined per Apache ActiveMQ Artemis
server.</p>
<pre><code>&lt;cluster-connections&gt;
&lt;cluster-connection name=&quot;my-cluster&quot;&gt;
&lt;address&gt;jms&lt;/address&gt;
&lt;connector-ref&gt;netty-connector&lt;/connector-ref&gt;
&lt;check-period&gt;1000&lt;/check-period&gt;
&lt;connection-ttl&gt;5000&lt;/connection-ttl&gt;
&lt;min-large-message-size&gt;50000&lt;/min-large-message-size&gt;
&lt;call-timeout&gt;5000&lt;/call-timeout&gt;
&lt;retry-interval&gt;500&lt;/retry-interval&gt;
&lt;retry-interval-multiplier&gt;1.0&lt;/retry-interval-multiplier&gt;
&lt;max-retry-interval&gt;5000&lt;/max-retry-interval&gt;
&lt;initial-connect-attempts&gt;-1&lt;/initial-connect-attempts&gt;
&lt;reconnect-attempts&gt;-1&lt;/reconnect-attempts&gt;
&lt;use-duplicate-detection&gt;true&lt;/use-duplicate-detection&gt;
&lt;message-load-balancing&gt;ON_DEMAND&lt;/message-load-balancing&gt;
&lt;max-hops&gt;1&lt;/max-hops&gt;
&lt;confirmation-window-size&gt;32000&lt;/confirmation-window-size&gt;
&lt;call-failover-timeout&gt;30000&lt;/call-failover-timeout&gt;
&lt;notification-interval&gt;1000&lt;/notification-interval&gt;
&lt;notification-attempts&gt;2&lt;/notification-attempts&gt;
&lt;discovery-group-ref discovery-group-name=&quot;my-discovery-group&quot;/&gt;
&lt;/cluster-connection&gt;
&lt;/cluster-connections&gt;
</code></pre><p>In the above cluster connection all parameters have been explicitly
specified. The following shows all the available configuration options</p>
<ul>
<li><p><code>address</code> Each cluster connection only applies to addresses that
match the specified address field. An address is matched on the
cluster connection when it begins with the string specified in this
field. The address field on a cluster connection also supports comma
separated lists and an exclude syntax &apos;!&apos;. To prevent an address
from being matched on this cluster connection, prepend a cluster
connection address string with &apos;!&apos;.</p>
<p>In the case shown above the cluster connection will load balance
messages sent to addresses that start with <code>jms</code>. This cluster
connection, will, in effect apply to all JMS queues and topics since
they map to core queues that start with the substring &quot;jms&quot;.</p>
<p>The address can be any value and you can have many cluster
connections with different values of <code>address</code>, simultaneously
balancing messages for those addresses, potentially to different
clusters of servers. By having multiple cluster connections on
different addresses a single Apache ActiveMQ Artemis Server can effectively take
part in multiple clusters simultaneously.</p>
<p>Be careful not to have multiple cluster connections with overlapping
values of <code>address</code>, e.g. &quot;europe&quot; and &quot;europe.news&quot; since this
could result in the same messages being distributed between more
than one cluster connection, possibly resulting in duplicate
deliveries.</p>
<p>Examples:</p>
<ul>
<li>&apos;jms.eu&apos;
matches all addresses starting with &apos;jms.eu&apos;</li>
<li>&apos;!jms.eu&apos;
matches all address except for those starting with &apos;jms.eu&apos;</li>
<li>&apos;jms.eu.uk,jms.eu.de&apos;
matches all addresses starting with either &apos;jms.eu.uk&apos; or
&apos;jms.eu.de&apos;</li>
<li>&apos;jms.eu,!jms.eu.uk&apos;
matches all addresses starting with &apos;jms.eu&apos; but not those
starting with &apos;jms.eu.uk&apos;</li>
</ul>
<p>Notes:</p>
<ul>
<li>Address exclusion will always takes precedence over address
inclusion.</li>
<li>Address matching on cluster connections does not support
wild-card matching.</li>
</ul>
<p>This parameter is mandatory.</p>
</li>
<li><p><code>connector-ref</code>. This is the connector which will be sent to other
nodes in the cluster so they have the correct cluster topology.</p>
<p>This parameter is mandatory.</p>
</li>
<li><p><code>check-period</code>. The period (in milliseconds) used to check if the
cluster connection has failed to receive pings from another server.
Default is 30000.</p>
</li>
<li><p><code>connection-ttl</code>. This is how long a cluster connection should stay
alive if it stops receiving messages from a specific node in the
cluster. Default is 60000.</p>
</li>
<li><p><code>min-large-message-size</code>. If the message size (in bytes) is larger
than this value then it will be split into multiple segments when
sent over the network to other cluster members. Default is 102400.</p>
</li>
<li><p><code>call-timeout</code>. When a packet is sent via a cluster connection and
is a blocking call, i.e. for acknowledgements, this is how long it
will wait (in milliseconds) for the reply before throwing an
exception. Default is 30000.</p>
</li>
<li><p><code>retry-interval</code>. We mentioned before that, internally, cluster
connections cause bridges to be created between the nodes of the
cluster. If the cluster connection is created and the target node
has not been started, or say, is being rebooted, then the cluster
connections from other nodes will retry connecting to the target
until it comes back up, in the same way as a bridge does.</p>
<p>This parameter determines the interval in milliseconds between retry
attempts. It has the same meaning as the <code>retry-interval</code> on a
bridge (as described in <a href="core-bridges.html">Core Bridges</a>).</p>
<p>This parameter is optional and its default value is <code>500</code>
milliseconds.</p>
</li>
<li><p><code>retry-interval-multiplier</code>. This is a multiplier used to increase
the <code>retry-interval</code> after each reconnect attempt, default is 1.</p>
</li>
<li><p><code>max-retry-interval</code>. The maximum delay (in milliseconds) for
retries. Default is 2000.</p>
</li>
<li><p><code>initial-connect-attempts</code>. The number of times the system will try
to connect a node in the cluster initially. If the max-retry is
achieved this node will be considered permanently down and the
system will not route messages to this node. Default is -1 (infinite
retries).</p>
</li>
<li><p><code>reconnect-attempts</code>. The number of times the system will try to
reconnect to a node in the cluster. If the max-retry is achieved
this node will be considered permanently down and the system will
stop routing messages to this node. Default is -1 (infinite
retries).</p>
</li>
<li><p><code>use-duplicate-detection</code>. Internally cluster connections use
bridges to link the nodes, and bridges can be configured to add a
duplicate id property in each message that is forwarded. If the
target node of the bridge crashes and then recovers, messages might
be resent from the source node. By enabling duplicate detection any
duplicate messages will be filtered out and ignored on receipt at
the target node.</p>
<p>This parameter has the same meaning as <code>use-duplicate-detection</code> on
a bridge. For more information on duplicate detection, please see <a href="duplicate-detection.html">Duplicate Detection</a>.
Default is true.</p>
</li>
<li><p><code>message-load-balancing</code>. This parameter determines if/how
messages will be distributed between other nodes of the cluster.
It can be one of three values - <code>OFF</code>, <code>STRICT</code>, or <code>ON_DEMAND</code>
(default). This parameter replaces the deprecated
<code>forward-when-no-consumers</code> parameter.</p>
<p>If this is set to <code>OFF</code> then messages will never be forwarded to
another node in the cluster</p>
<p>If this is set to <code>STRICT</code> then each incoming message will be round
robin&apos;d even though the same queues on the other nodes of the
cluster may have no consumers at all, or they may have consumers
that have non matching message filters (selectors). Note that
Apache ActiveMQ Artemis will <em>not</em> forward messages to other nodes
if there are no <em>queues</em> of the same name on the other nodes, even
if this parameter is set to <code>STRICT</code>. Using <code>STRICT</code> is like setting
the legacy <code>forward-when-no-consumers</code> parameter to <code>true</code>.</p>
<p>If this is set to <code>ON_DEMAND</code> then Apache ActiveMQ Artemis will only
forward messages to other nodes of the cluster if the address to which
they are being forwarded has queues which have consumers, and if those
consumers have message filters (selectors) at least one of those
selectors must match the message. Using <code>ON_DEMAND</code> is like setting
the legacy <code>forward-when-no-consumers</code> parameter to <code>false</code>.</p>
<p>Default is <code>ON_DEMAND</code>.</p>
</li>
<li><p><code>max-hops</code>. When a cluster connection decides the set of nodes to
which it might load balance a message, those nodes do not have to be
directly connected to it via a cluster connection. Apache ActiveMQ Artemis can be
configured to also load balance messages to nodes which might be
connected to it only indirectly with other Apache ActiveMQ Artemis servers as
intermediates in a chain.</p>
<p>This allows Apache ActiveMQ Artemis to be configured in more complex topologies and
still provide message load balancing. We&apos;ll discuss this more later
in this chapter.</p>
<p>The default value for this parameter is <code>1</code>, which means messages
are only load balanced to other Apache ActiveMQ Artemis serves which are directly
connected to this server. This parameter is optional.</p>
</li>
<li><p><code>confirmation-window-size</code>. The size (in bytes) of the window used
for sending confirmations from the server connected to. So once the
server has received <code>confirmation-window-size</code> bytes it notifies its
client, default is 1048576. A value of -1 means no window.</p>
</li>
<li><p><code>producer-window-size</code>. The size for producer flow control over cluster connection.
it&apos;s by default disabled through the cluster connection bridge but you may want
to set a value if you are using really large messages in cluster. A value of -1 means no window.</p>
</li>
<li><p><code>call-failover-timeout</code>. Similar to <code>call-timeout</code> but used when a
call is made during a failover attempt. Default is -1 (no timeout).</p>
</li>
<li><p><code>notification-interval</code>. How often (in milliseconds) the cluster
connection should broadcast itself when attaching to the cluster.
Default is 1000.</p>
</li>
<li><p><code>notification-attempts</code>. How many times the cluster connection
should broadcast itself when connecting to the cluster. Default is
2.</p>
</li>
<li><p><code>discovery-group-ref</code>. This parameter determines which discovery
group is used to obtain the list of other servers in the cluster
that this cluster connection will make connections to.</p>
</li>
</ul>
<p>Alternatively if you would like your cluster connections to use a static
list of servers for discovery then you can do it like this.</p>
<pre><code>&lt;cluster-connection name=&quot;my-cluster&quot;&gt;
...
&lt;static-connectors&gt;
&lt;connector-ref&gt;server0-connector&lt;/connector-ref&gt;
&lt;connector-ref&gt;server1-connector&lt;/connector-ref&gt;
&lt;/static-connectors&gt;
&lt;/cluster-connection&gt;
</code></pre><p>Here we have defined 2 servers that we know for sure will that at least
one will be available. There may be many more servers in the cluster but
these will; be discovered via one of these connectors once an initial
connection has been made.</p>
<h3 id="cluster-user-credentials">Cluster User Credentials</h3>
<p>When creating connections between nodes of a cluster to form a cluster
connection, Apache ActiveMQ Artemis uses a cluster user and cluster password which is
defined in <code>broker.xml</code>:</p>
<pre><code>&lt;cluster-user&gt;ACTIVEMQ.CLUSTER.ADMIN.USER&lt;/cluster-user&gt;
&lt;cluster-password&gt;CHANGE ME!!&lt;/cluster-password&gt;
</code></pre><blockquote>
<p><strong>Warning</strong></p>
<p>It is imperative that these values are changed from their default, or
remote clients will be able to make connections to the server using
the default values. If they are not changed from the default, Apache ActiveMQ Artemis
will detect this and pester you with a warning on every start-up.</p>
</blockquote>
<h2 id="client-side-load-balancing">Client-Side Load balancing</h2>
<p>With Apache ActiveMQ Artemis client-side load balancing, subsequent sessions created
using a single session factory can be connected to different nodes of
the cluster. This allows sessions to spread smoothly across the nodes of
a cluster and not be &quot;clumped&quot; on any particular node.</p>
<p>The load balancing policy to be used by the client factory is
configurable. Apache ActiveMQ Artemis provides four out-of-the-box load balancing
policies, and you can also implement your own and use that.</p>
<p>The out-of-the-box policies are</p>
<ul>
<li><p>Round Robin. With this policy the first node is chosen randomly then
each subsequent node is chosen sequentially in the same order.</p>
<p>For example nodes might be chosen in the order B, C, D, A, B, C, D,
A, B or D, A, B, C, D, A, B, C, D or C, D, A, B, C, D, A, B, C.</p>
<p>Use
<code>org.apache.activemq.artemis.api.core.client.loadbalance.RoundRobinConnectionLoadBalancingPolicy</code>
as the <code>&lt;connection-load-balancing-policy-class-name&gt;</code>.</p>
</li>
<li><p>Random. With this policy each node is chosen randomly.</p>
<p>Use
<code>org.apache.activemq.artemis.api.core.client.loadbalance.RandomConnectionLoadBalancingPolicy</code>
as the <code>&lt;connection-load-balancing-policy-class-name&gt;</code>.</p>
</li>
<li><p>Random Sticky. With this policy the first node is chosen randomly
and then re-used for subsequent connections.</p>
<p>Use
<code>org.apache.activemq.artemis.api.core.client.loadbalance.RandomStickyConnectionLoadBalancingPolicy</code>
as the <code>&lt;connection-load-balancing-policy-class-name&gt;</code>.</p>
</li>
<li><p>First Element. With this policy the &quot;first&quot; (i.e. 0th) node is
always returned.</p>
<p>Use
<code>org.apache.activemq.artemis.api.core.client.loadbalance.FirstElementConnectionLoadBalancingPolicy</code>
as the <code>&lt;connection-load-balancing-policy-class-name&gt;</code>.</p>
</li>
</ul>
<p>You can also implement your own policy by implementing the interface
<code>org.apache.activemq.artemis.api.core.client.loadbalance.ConnectionLoadBalancingPolicy</code></p>
<p>Specifying which load balancing policy to use differs whether you are
using JMS or the core API. If you don&apos;t specify a policy then the
default will be used which is
<code>org.apache.activemq.artemis.api.core.client.loadbalance.RoundRobinConnectionLoadBalancingPolicy</code>.</p>
<p>If you&apos;re using JMS and you&apos;re using JNDI on the client to look up your
JMS connection factory instances then you can specify these parameters
in the JNDI context environment in, e.g. <code>jndi.properties</code>, to specify
the load balancing policy directly:</p>
<pre><code>java.naming.factory.initial=org.apache.activemq.artemis.jndi.ActiveMQInitialContextFactory
connection.myConnectionFactory=tcp://localhost:61616?loadBalancingPolicyClassName=org.apache.activemq.artemis.api.core.client.loadbalance.RandomConnectionLoadBalancingPolicy
</code></pre><p>The above example would instantiate a JMS connection factory that uses
the random connection load balancing policy.</p>
<p>If you&apos;re using JMS but you&apos;re instantiating your connection factory
directly on the client side then you can set the load balancing policy
using the setter on the <code>ActiveMQConnectionFactory</code> before using it:</p>
<pre><code class="lang-java">ConnectionFactory jmsConnectionFactory = ActiveMQJMSClient.createConnectionFactory(...);
jmsConnectionFactory.setLoadBalancingPolicyClassName(<span class="hljs-string">&quot;com.acme.MyLoadBalancingPolicy&quot;</span>);
</code></pre>
<p>If you&apos;re using the core API, you can set the load balancing policy
directly on the <code>ServerLocator</code> instance you are using:</p>
<pre><code class="lang-java">ServerLocator locator = ActiveMQClient.createServerLocatorWithHA(server1, server2);
locator.setLoadBalancingPolicyClassName(<span class="hljs-string">&quot;com.acme.MyLoadBalancingPolicy&quot;</span>);
</code></pre>
<p>The set of servers over which the factory load balances can be
determined in one of two ways:</p>
<ul>
<li><p>Specifying servers explicitly</p>
</li>
<li><p>Using discovery.</p>
</li>
</ul>
<h2 id="specifying-members-of-a-cluster-explicitly">Specifying Members of a Cluster Explicitly</h2>
<p>Sometimes you want to explicitly define a cluster more explicitly, that
is control which server connect to each other in the cluster. This is
typically used to form non symmetrical clusters such as chain cluster or
ring clusters. This can only be done using a static list of connectors
and is configured as follows:</p>
<pre><code>&lt;cluster-connection name=&quot;my-cluster&quot;&gt;
&lt;address&gt;jms&lt;/address&gt;
&lt;connector-ref&gt;netty-connector&lt;/connector-ref&gt;
&lt;retry-interval&gt;500&lt;/retry-interval&gt;
&lt;use-duplicate-detection&gt;true&lt;/use-duplicate-detection&gt;
&lt;message-load-balancing&gt;STRICT&lt;/message-load-balancing&gt;
&lt;max-hops&gt;1&lt;/max-hops&gt;
&lt;static-connectors allow-direct-connections-only=&quot;true&quot;&gt;
&lt;connector-ref&gt;server1-connector&lt;/connector-ref&gt;
&lt;/static-connectors&gt;
&lt;/cluster-connection&gt;
</code></pre><p>In this example we have set the attribute
<code>allow-direct-connections-only</code> which means that the only server that
this server can create a cluster connection to is server1-connector.
This means you can explicitly create any cluster topology you want.</p>
<h2 id="message-redistribution">Message Redistribution</h2>
<p>Another important part of clustering is message redistribution. Earlier
we learned how server side message load balancing round robins messages
across the cluster. If <code>message-load-balancing</code> is <code>OFF</code> or <code>ON_DEMAND</code>
then messages won&apos;t be forwarded to nodes which don&apos;t have matching
consumers. This is great and ensures that messages aren&apos;t moved to a
queue which has no consumers to consume them. However, there is a
situation it doesn&apos;t solve: What happens if the consumers on a queue
close after the messages have been sent to the node? If there are no
consumers on the queue the message won&apos;t get consumed and we have a
<em>starvation</em> situation.</p>
<p>This is where message redistribution comes in. With message
redistribution Apache ActiveMQ Artemis can be configured to automatically
<em>redistribute</em> messages from queues which have no consumers back to
other nodes in the cluster which do have matching consumers. To enable
this functionality <code>message-load-balancing</code> must be <code>ON_DEMAND</code>.</p>
<p>Message redistribution can be configured to kick in immediately after
the last consumer on a queue is closed, or to wait a configurable delay
after the last consumer on a queue is closed before redistributing. By
default message redistribution is disabled.</p>
<p>Message redistribution can be configured on a per address basis, by
specifying the redistribution delay in the address settings, for more
information on configuring address settings, please see <a href="queue-attributes.html">Queue Attributes</a>.</p>
<p>Here&apos;s an address settings snippet from <code>broker.xml</code>
showing how message redistribution is enabled for a set of queues:</p>
<pre><code>&lt;address-settings&gt;
&lt;address-setting match=&quot;jms.#&quot;&gt;
&lt;redistribution-delay&gt;0&lt;/redistribution-delay&gt;
&lt;/address-setting&gt;
&lt;/address-settings&gt;
</code></pre><p>The above <code>address-settings</code> block would set a <code>redistribution-delay</code> of
<code>0</code> for any queue which is bound to an address that starts with &quot;jms.&quot;.
All JMS queues and topic subscriptions are bound to addresses that start
with &quot;jms.&quot;, so the above would enable instant (no delay) redistribution
for all JMS queues and topic subscriptions.</p>
<p>The attribute <code>match</code> can be an exact match or it can be a string that
conforms to the Apache ActiveMQ Artemis wildcard syntax (described in <a href="wildcard-syntax.html">Wildcard Syntax</a>).</p>
<p>The element <code>redistribution-delay</code> defines the delay in milliseconds
after the last consumer is closed on a queue before redistributing
messages from that queue to other nodes of the cluster which do have
matching consumers. A delay of zero means the messages will be
immediately redistributed. A value of <code>-1</code> signifies that messages will
never be redistributed. The default value is <code>-1</code>.</p>
<p>It often makes sense to introduce a delay before redistributing as it&apos;s
a common case that a consumer closes but another one quickly is created
on the same queue, in such a case you probably don&apos;t want to
redistribute immediately since the new consumer will arrive shortly.</p>
<h2 id="cluster-topologies">Cluster topologies</h2>
<p>Apache ActiveMQ Artemis clusters can be connected together in many different
topologies, let&apos;s consider the two most common ones here</p>
<h3 id="symmetric-cluster">Symmetric cluster</h3>
<p>A symmetric cluster is probably the most common cluster topology.</p>
<p>With a symmetric cluster every node in the cluster is connected to every
other node in the cluster. In other words every node in the cluster is
no more than one hop away from every other node.</p>
<p>To form a symmetric cluster every node in the cluster defines a cluster
connection with the attribute <code>max-hops</code> set to <code>1</code>. Typically the
cluster connection will use server discovery in order to know what other
servers in the cluster it should connect to, although it is possible to
explicitly define each target server too in the cluster connection if,
for example, UDP is not available on your network.</p>
<p>With a symmetric cluster each node knows about all the queues that exist
on all the other nodes and what consumers they have. With this knowledge
it can determine how to load balance and redistribute messages around
the nodes.</p>
<p>Don&apos;t forget <a href="#copy-warning">this warning</a> when creating a symmetric
cluster.</p>
<h3 id="chain-cluster">Chain cluster</h3>
<p>With a chain cluster, each node in the cluster is not connected to every
node in the cluster directly, instead the nodes form a chain with a node
on each end of the chain and all other nodes just connecting to the
previous and next nodes in the chain.</p>
<p>An example of this would be a three node chain consisting of nodes A, B
and C. Node A is hosted in one network and has many producer clients
connected to it sending order messages. Due to corporate policy, the
order consumer clients need to be hosted in a different network, and
that network is only accessible via a third network. In this setup node
B acts as a mediator with no producers or consumers on it. Any messages
arriving on node A will be forwarded to node B, which will in turn
forward them to node C where they can get consumed. Node A does not need
to directly connect to C, but all the nodes can still act as a part of
the cluster.</p>
<p>To set up a cluster in this way, node A would define a cluster
connection that connects to node B, and node B would define a cluster
connection that connects to node C. In this case we only want cluster
connections in one direction since we&apos;re only moving messages from node
A-&gt;B-&gt;C and never from C-&gt;B-&gt;A.</p>
<p>For this topology we would set <code>max-hops</code> to <code>2</code>. With a value of <code>2</code>
the knowledge of what queues and consumers that exist on node C would be
propagated from node C to node B to node A. Node A would then know to
distribute messages to node B when they arrive, even though node B has
no consumers itself, it would know that a further hop away is node C
which does have consumers.</p>
<h3 id="scaling-down">Scaling Down</h3>
<p>Apache ActiveMQ Artemis supports scaling down a cluster with no message loss (even for
non-durable messages). This is especially useful in certain environments
(e.g. the cloud) where the size of a cluster may change relatively
frequently. When scaling up a cluster (i.e. adding nodes) there is no
risk of message loss, but when scaling down a cluster (i.e. removing
nodes) the messages on those nodes would be lost unless the broker sent
them to another node in the cluster. Apache ActiveMQ Artemis can be configured to do
just that.</p>
<p>The simplest way to enable this behavior is to set <code>scale-down</code> to
<code>true</code>. If the server is clustered and <code>scale-down</code> is <code>true</code> then when
the server is shutdown gracefully (i.e. stopped without crashing) it
will find another node in the cluster and send <em>all</em> of its messages
(both durable and non-durable) to that node. The messages are processed
in order and go to the <em>back</em> of the respective queues on the other node
(just as if the messages were sent from an external client for the first
time).</p>
<p>If more control over where the messages go is required then specify
<code>scale-down-group-name</code>. Messages will only be sent to another node in
the cluster that uses the same <code>scale-down-group-name</code> as the server
being shutdown.</p>
<blockquote>
<p><strong>Warning</strong></p>
<p>If cluster nodes are grouped together with different
<code>scale-down-group-name</code> values beware. If all the nodes in a single
group are shut down then the messages from that node/group will be
lost.</p>
</blockquote>
<p>If the server is using multiple <code>cluster-connection</code> then use
<code>scale-down-clustername</code> to identify the name of the
<code>cluster-connection</code> which should be used for scaling down.</p>
</section>
</div>
</div>
</div>
<a href="./duplicate-detection.html" class="navigation navigation-prev " aria-label="Previous page: Duplicate Message Detection"><i class="fa fa-angle-left"></i></a>
<a href="./ha.html" class="navigation navigation-next " aria-label="Next page: High Availability and Failover"><i class="fa fa-angle-right"></i></a>
</div>
</div>
<script src="gitbook/app.js"></script>
<script>
require(["gitbook"], function(gitbook) {
var config = {"fontSettings":{"theme":null,"family":"sans","size":2}};
gitbook.start(config);
});
</script>
</body>
</html>