[jboss-cvs] jboss-seam/doc/reference/en/modules ...
Gavin King
gavin.king at jboss.com
Mon Feb 26 23:18:40 EST 2007
User: gavin
Date: 07/02/26 23:18:40
Added: doc/reference/en/modules cache.xml
Log:
JBSEAM-936
Revision Changes Path
1.1 date: 2007/02/27 04:18:40; author: gavin; state: Exp;jboss-seam/doc/reference/en/modules/cache.xml
Index: cache.xml
===================================================================
<chapter id="cache">
<title>Caching</title>
<para>
In almost all enterprise applications, the database is the primary
bottleneck, and the least scalable tier of the runtime environment.
People from a PHP/Ruby environment will try to tell you that so-called
"shared nothing" architectures scale well. While that may be literally
true, I don't know of many interesting multi-user applications which
can be implemented with no sharing of resources between different
nodes of the cluster. What these silly people are really thinking of
is a "share nothing except for the database" architecture. Of course,
sharing the database is the primary problem with scaling a multi-user
application—so the claim that this architecture is highly scalable
is absurd, and tells you a lot about the kind of applications that these
folks spend most of their time working on.
</para>
<para>
Almost anything we can possibly do to share the database <emphasis>less
often</emphasis> is worth doing.
</para>
<para>
This calls for a cache. Well, not just one cache. A well designed Seam
application will feature a rich, multi-layered caching strategy that
impacts every layer of the application:
</para>
<itemizedlist>
<listitem>
<para>
The database, of course, has its own cache. This is
super-important, but can't scale like a cache in the
application tier.
</para>
</listitem>
<listitem>
<para>
Your ORM solution (Hibernate, or some other JPA implementation)
has a second-level cache of data from the database. This is
a very powerful capability, but is often misused. In a clustered
environment, keeping the data in the cache transactionally
consistent across the whole cluster, and with the database,
is quite expensive. It makes most sense for data which is shared
between many users, and is updated rarely. In traditional
stateless architectures, people often try to use the second-level
cache for conversational state. This is always bad, and is
especially wrong in Seam.
</para>
</listitem>
<listitem>
<para>
The Seam conversation context is a cache of conversational state.
Components you put into the conversation context can hold and cache
state relating to the current user interaction.
</para>
</listitem>
<listitem>
<para>
In particular, the
Seam-managed persistence context (or an extended EJB container-managed
persistence context associated with a conversation-scoped stateful
session bean) acts as a cache of data that has been read in the
current conversation. This cache tends to have a pretty high
hitrate! Seam optimizes the replication of Seam-managed persistence
contexts in a clustered environment, and there is no requirement for
transactional consistency with the database (optimistic locking is
sufficient) so you don't need to worry too much about the performance
implications of this cache, unless you read thousands of objects into
a single persistence context.
</para>
</listitem>
<listitem>
<para>
The application can cache non-transactional state in the Seam
application context. State kept in the application context is
of course not visible to other nodes in the cluster.
</para>
</listitem>
<listitem>
<para>
The application can cache transactional state using the Seam
<literal>pojoCache</literal> component, which integrates
JBossCache into the Seam environment. This state will be visible
to other nodes if you run JBoss cache in a clustered mode.
</para>
</listitem>
<listitem>
<para>
Finally, Seam lets you cache rendered fragments of a JSF page. Unlike
the ORM second-level cache, this cache is not automatically invalidated
when data changes, so you need to write application code to perform
explicit invalidation, or set appropriate expiration policies.
</para>
</listitem>
</itemizedlist>
<para>
For more information about the second-level cache, you'll need to refer to
the documentation of your ORM solution, since this is an extremely complex
topic. In this section we'll discuss the use of JBossCache directly, via
the <literal>pojoCache</literal> component, or as the page fragment cache,
via the <literal><s:cache></literal> control.
</para>
<section>
<title>Using JBossCache in Seam</title>
<para>
The built-in <literal>pojoCache</literal> component manages an instance
of <literal>org.jboss.cache.aop.PojoCache</literal>. You can safely put
any immutable Java object in the cache, and it will be replicated across
the cluster (assuming that replication is enabled). If you want to
keep mutable objects in the cache, you'll need to run the JBossCache
bytecode preprocessor to ensure that changes to the objects will be
automatically detected and replicated.
</para>
<para>
To use <literal>pojoCache</literal>, all you need to do is put the
JBossCache jars in the classpath, and provide a resource named
<literal>treecache.xml</literal> with an appropriate cache configuration.
JBossCache has many scary and confusing configuration settings, so
we won't discuss them here. Please refer to the JBossCache documentation
for more information.
</para>
<para>
For an EAR depoyment of Seam, we recommend that the JBossCache jars and
configuration go directly into the EAR. Make sure you declare the jars
in <literal>application.xml</literal>.
</para>
<para>
Now you can inject the cache into any Seam component:
</para>
<programlisting><![CDATA[@Name("chatroom")
public class Chatroom {
@In PojoCache pojoCache;
public void join(String username) {
try
{
Set<String> userList = (Set<String>) pojoCache.get("chatroom", "userList");
if (users==null)
{
userList = new HashSet<String>();
pojoCache.put("chatroom", "users", username);
}
}
catch (CacheException ce)
{
throw new RuntimeException(ce);
}
}
}]]></programlisting>
<para>
If you want to have multiple JBossCache configurations in your application,
use <literal>components.xml</literal>:
</para>
<programlisting><![CDATA[<core:pojo-cache name="myCache" cfg-resource-name="myown/cache.xml"/>]]></programlisting>
</section>
<section>
<title>Page fragment caching</title>
<para>
The most interesting user of JBossCache is the <literal><s:cache></literal>
tag, Seam's solution to the problem of page fragment caching in JSF.
<literal><s:cache></literal> uses <literal>pojoCache</literal> internally,
so you need to follow the steps listed above before you can use it. (Put the jars
in the EAR, wade through the scary configuration options, etc.)
</para>
<para>
<literal><s:cache></literal> is used for caching some rendered content which
changes rarely. For example, the welcome page of our blog displays the recent
blog entries:
</para>
<programlisting><![CDATA[<s:cache key="recentEntries-#{blog.id}" region="welcomePageFragments">
<h:dataTable value="#{blog.recentEntries}" var="blogEntry">
<h:column>
<h3>#{blogEntry.title}</h3>
<div>
<s:formattedText value="#{blogEntry.body}"/>
</div>
</h:column>
</h:dataTable>
</s:cache>]]></programlisting>
<para>
The <literal>key</literal> let's you have multiple cached versions of each page
fragment. In this case, there is one cached version per blog. The
<literal>region</literal> determines the JBossCache node that all version will
be stored in. Different nodes may have different expiry policies. (That's the
stuff you set up using the aforementioned scary configuration options.)
</para>
<para>
Of course, the big problem with <literal><s:cache></literal> is that it
is too stupid to know when the underlying data changes (for example, when the
blogger posts a new entry). So you need to evict the cached fragment manually:
</para>
<programlisting><![CDATA[public void post() {
...
entityManager.persist(blogEntry);
pojoCache.remove("welcomePageFragments", "recentEntries-" + blog.getId() );
}]]></programlisting>
<para>
Alternatively, if it is not critical that changes are immediately visible to the
user, you could set a short expiry time on the JbossCache node.
</para>
</section>
</chapter>
More information about the jboss-cvs-commits
mailing list