<?xml version="1.0" encoding="ISO-8859-1"?>
<!DOCTYPE document [
  <!ENTITY project SYSTEM "project.xml">
]>
<document url="workers.html">

  &project;
<copyright>
   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.
</copyright>
<properties>
<title>Workers HowTo</title>
<author email="hgomez@apache.org">Henri Gomez</author>
<author email="shachor@il.ibm.com">Gal Shachor</author>
<author email="mturk@apache.org">Mladen Turk</author>
<date>$Date$</date>
</properties>
<body>
<section name="Introduction">
<p>
A Tomcat worker is a Tomcat instance that is waiting to execute servlets on behalf of some web server. 
For example, we can have a web server such as the Apache HTTP Server forwarding servlet requests to a 
Tomcat process (the worker) running behind it.
</p>
<p>
The scenario described above is a very simple one; 
in fact one can configure multiple Tomcat workers to serve servlets on 
behalf of a certain web server. 
The reasons for such configuration can be:
</p>
<ul>
<li>
We want different contexts to be served by different Tomcat workers to provide a 
development environment where all the developers share the same web server but own a Tomcat worker of their own.
</li>
<li>
We want different virtual hosts served by different Tomcat processes to provide a 
clear separation between sites belonging to different companies.
</li>
<li>
We want to provide load balancing, meaning run multiple Tomcat workers each on a 
machine of its own and distribute the requests between them.
</li>
</ul>

<p>
There are probably more reasons for having multiple workers but I guess that this list is enough...
Tomcat workers are defined in a properties file dubbed workers.properties and this tutorial 
explains how to work with it.
</p>

<p>
This document was originally part of <b>Tomcat: A Minimalistic User's Guide</b> written by Gal Shachor, 
but has been split off for organisational reasons. 
</p>
</section>

<section name="Defining Workers">
<p>
Defining workers to the Tomcat web server plugin can be done using a properties file 
(a sample file named workers.properties is available in the conf/ directory).
</p>

<p>
the file contains entries of the following form:
</p>

<p>
<b>worker.list</b>=&lt;a comma separated list of worker names&gt;
</p>

<source>
  # the list of workers
  worker.list= worker1, worker2
</source>

<p>
When starting up, the web server plugin will instantiate the workers whose name appears in the 
<b>worker.list</b> property, these are also the workers to whom you can map requests. The directive can be used multiple times.
</p>

<subsection name="Worker Types">
<p>
Each named worker should also have a few entries to provide additional information on his behalf.
This information includes the worker's type and other related worker information. 
Currently the following worker types that exists are (JK 1.2.5):
</p>

<table>
  <tr><th>Type</th><th>Description</th></tr>
  <tr><td>ajp13</td><td>This worker knows how to forward requests to out-of-process Tomcat workers using the ajp13 protocol.</td></tr>
  <tr><td>lb</td><td>This is a load balancing worker; it knows how to provide flexible load balancing with a certain level of fault-tolerance.</td></tr>
  <tr><td>status</td><td>This is a status worker for managing load balancers.</td></tr>
  <tr><td>ajp12</td><td>This worker knows how to forward requests to out-of-process Tomcat workers using the ajp12 protocol. It is deprecated</td></tr>
  <tr><td>ajp14</td><td>This worker knows how to forward requests to out-of-process Tomcat workers using the ajp14 protocol. It is experimental</td></tr>
</table>

<p>
Defining workers of a certain type should be done with the following property format:
</p>

<p>
<b>worker</b>.<b>worker name</b>.<b>type</b>=&lt;worker type&gt;
Where worker name is the name assigned to the worker and the worker type is one of the four types defined 
in the table (a worker name may only contain any space the characters [a-zA-Z0-9\-_]).
</p>

<source>
  # Defines a worker named "local" that uses the ajp12 protocol to forward requests to a Tomcat process.
  worker.local.type=ajp12
  # Defines a worker named "remote" that uses the ajp13 protocol to forward requests to a Tomcat process.
  worker.remote.type=ajp13
  # Defines a worker named "loadbalancer" that loadbalances several Tomcat processes transparently.
  worker.loadbalancer.type=lb
</source>

</subsection>

</section>

<section name="Setting Worker Properties">
<p>
After defining the workers you can also specify properties for them. 
Properties can be specified in the following manner:
</p>

<p>
worker.&lt;worker name&gt;.&lt;property&gt;=&lt;property value&gt;
</p>

Each worker has a set of properties that you can set as specified in the following subsections:

<subsection name="ajp12 Worker Properties">
<p><warn>
The worker type <b>ajp12</b> has been <b>deprecated</b> and you should use instead 
<b>ajp13</b> instead.
</warn></p> 
<p>
The ajp12 typed workers forward requests to out-of-process Tomcat workers 
using the ajp12 protocol over TCP/IP sockets. It does not use persistent connections.
</p>

<p>
The ajp12 worker properties are:
</p>

<p>
<b>host</b> property sets the host where the Tomcat worker is listening for ajp12 requests.
</p>

<p>
<b>port</b> property sets the port where the Tomcat worker is listening for ajp12 requests
</p>

<p>
<b>lbfactor</b> property is used when working with a load balancer worker, this is the load balancing factor for the worker.
We'll see more on this in the <a href="../common_howto/loadbalancers.html">load balancer worker</a> section.
</p>

<source>
  # worker "worker1" will talk to Tomcat listening on machine www.x.com at port 8007 using 2 lb factor
  worker.worker1.type=ajp12
  worker.worker1.host=www.x.com
  worker.worker1.port=8007
  worker.worker1.lbfactor=2
</source>

<p>
Note: The default port for ajp12 is 8007
</p>

</subsection>

<subsection name="ajp13 Worker Properties">
<p>
The ajp13 typed workers forward requests to out-of-process Tomcat workers using the ajp13 protocol over TCP/IP sockets.
The main difference between ajp12 and ajp13 are that:
<ul>
<li>
ajp13 is a more binary protocol and it tries to compress some of the request data by coding 
frequently used strings as small integers.
</li>
<li>
ajp13 reuses open sockets and leaves them open for future requests (remember when you've got a Firewall between your 
web server and Tomcat).
</li>
<li>
ajp13 has special treatment for SSL information so that the container can implement 
SSL related methods such as isSecure().
</li>
</ul>

</p>

<p>
You should note that ajp13 is the recommended protocol to connect to Tomcat.
</p>


<source>
  # worker "worker2" will talk to Tomcat listening on machine www2.x.com at port 8009 using 3 lb factor
  worker.worker2.type=ajp13
  worker.worker2.host=www2.x.com
  worker.worker2.port=8009
  worker.worker2.lbfactor=3
  # worker "worker2" uses connections, which will stay no more than 10mn in the connection pool
  worker.worker2.connection_pool_timeout=600
  # worker "worker2" ask operating system to send KEEP-ALIVE signal on the connection
  worker.worker2.socket_keepalive=1
  # mount can be used as an alternative to the JkMount directive
  worker.worker2.mount=/contexta /contexta/* /contextb /contextb/*
</source>

<p>
Notes: In the ajp13 protocol, the default port is 8009
</p>

</subsection>

<subsection name="lb Worker Properties">
<p>
The load balancing worker does not really communicate with Tomcat workers.
Instead it is responsible for the management of several "real" workers. 
This management includes:
</p>

<ul>
<li>
Instantiating the workers in the web server.
</li>
<li>
Using the worker's load balancing factor, perform weighted round-robin load balancing where 
a higher lbfactor means stronger machine (that is going to handle proportionally more requests)
</li>
<li>
Keeping requests belonging to the same session executing on the same Tomcat worker
(session stickyness).
</li>
<li>
Identifying failed Tomcat workers, suspending requests to them and instead failing over on 
other workers managed by the lb worker.
</li>
</ul>

<p>
The overall result is that workers managed by the same lb worker are load balanced
(based on their lbfactor and current user session) and also fail over so a single
Tomcat process death will not "kill" the entire site.
The following table specifies some properties that the lb worker can accept:
<ul>
<li><b>balance_workers</b> is a comma separated list of workers that the load balancer need to manage.
As long as these workers should only be used via the load balancer worker,
there is no need to also put them into the worker.list property.
This directive can be used multiple times for the same load balancer.</li>
<li><b>sticky_session</b> specifies whether requests with SESSION ID's should be routed back to the same
Tomcat worker. Set sticky_session to <b>false</b> when Tomcat is using a Session Manager which
can persist session data across multiple instances of Tomcat. By default sticky_session is set to
<b>true</b>.</li>
</ul>
</p>

<source>
  # The worker balance1 while use "real" workers worker1 and worker2
  worker.balance1.balance_workers=worker1, worker2
</source>

</subsection>

<subsection name="Status Worker properties">
<p>
The status worker does not communicate with Tomcat.
Instead it is responsible for the load balancer management. 
</p>
<source>
  # Add the status worker to the worker list
  worker.list=jkstatus
  # Define a 'jkstatus' worker using status
  worker.jkstatus.type=status
</source>
<p>Next thing is to mount the requests to the jkstatus worker. For Apache
HTTP Servers use:</p>
<source>
  # Add the jkstatus mount point
  JkMount /jkmanager/* jkstatus 
</source>
<p>To obtain a higher level of security use the:</p>
<source>
  # Enable the JK manager access from localhost only
 &lt;Location /jkmanager/&gt;
    JkMount jkstatus
    Require ip 127.0.0.1
 &lt;/Location&gt;
</source>

</subsection>

<subsection name="Property file macros">
<p>
You can define "macros" in the property files. 
These macros let you define properties and later on use them while 
constructing other properties.
</p>

<source>
  # property example, like a network base address
  mynet=194.226.31
  # Using the above macro to simplify the address definitions
  # for a farm of workers.
  worker.node1.host=$(mynet).11
  worker.node2.host=$(mynet).12
  worker.node3.host=$(mynet).13
</source>

</subsection>

<subsection name="Hierarchical property configuration">
<p>
Workers can reference configurations of other workers.
If worker "x" references worker "y", then it inherits all
configuration parameters from "y", except for the ones
that have explicitly been set for "x".
</p>

<source>
  # worker toe defines some default settings
  worker.toe.type=ajp13
  worker.toe.socket_keepalive=true
  worker.toe.connect_timeout=10000
  worker.toe.recovery_options=7
  # workers tic and tac inherit those values
  worker.tic.reference=worker.toe
  worker.tac.reference=worker.toe
</source>

<p>
Please note, that the reference contains
the full prefix to the referenced configuration attributes,
not only the name of the referenced worker.
</p>

<p>
References can be nested with a maximum depth of 20.
Be careful to avoid loops!
</p>

<p>
Attributes which are allowed multiple times for a single worker
can not be merged from a worker and a reference. An attribute
is only inherited from a reference, if it is not already set
for the referring worker.
</p>

<p>
References are especially useful, when configuring load balancers.
Try to understand the following two stage references:
</p>

<source>
  # We only use one load balancer
  worker.list=lb
  # Let's define some defaults
  worker.basic.port=8009
  worker.basic.type=ajp13
  worker.basic.socket_keepalive=true
  worker.basic.connect_timeout=10000
  worker.basic.recovery_options=7
  # And we use them in two groups
  worker.lb1.domain=dom1
  worker.lb1.distance=0
  worker.lb1.reference=worker.basic
  worker.lb2.domain=dom2
  worker.lb2.distance=1
  worker.lb2.reference=worker.basic
  # Now we configure the load balancer
  worker.lb.type=lb
  worker.lb.method=B
  worker.lb.balanced_workers=w11,w12,w21,w22
  worker.w11.host=myhost11
  worker.w11.reference=worker.lb1
  worker.w12.host=myhost12
  worker.w12.reference=worker.lb1
  worker.w21.host=myhost21
  worker.w21.reference=worker.lb2
  worker.w22.host=myhost22
  worker.w22.reference=worker.lb2
</source>

</subsection>

</section>

<section name="A sample worker.properties">
<p>
Since coping with worker.properties on your own is not an easy thing to do, 
a sample worker.properties file is bundled along JK. 
</p>

<p>
You could also find here a sample workers.properties defining:
</p>

<ul>
<li>
Two ajp13 workers that use the host localhost and the ports 8109 and 8209
</li>
<li>
An lb worker that load balances over the two ajp13 workers
</li>
</ul>

<source>
  # Define 3 workers, 2 real workers using ajp13, and one being a load balancing worker 
  worker.list=node1, node2, lbworker
  # Set properties for node1 (ajp13)
  worker.node1.type=ajp13
  worker.node1.host=localhost
  worker.node1.port=8109
  worker.node1.connection_pool_timeout=600
  worker.node1.socket_keepalive=1
  # Set properties for node2 (ajp13)
  worker.node2.type=ajp13
  worker.node2.host=localhost
  worker.node2.port=8209
  worker.node2.connection_pool_timeout=600
  worker.node2.socket_keepalive=1
  # Set properties for lbworker which uses node1 and node2
  worker.lbworker.type=lb
  worker.lbworker.balance_workers=node1,node2
</source>

</section>
</body>
</document>