Develop Reference

Java - DBCP vs JNDI?

I am a newbie and I have created a few simlpe Java Swing applications. I was able to use apache commons DBCP to create a connection pool and access the datasource.
I have recently started to create java web based applciations using JSP and Servlets. I have learnt to use JNDI to access the datasource. I update the XML files and use InitialContext() and lookup("java:comp/env") and that is it!!!! I am using Apache Tomcat as my Servlet/JSP container.
1. But where is the DB connection pool created?
2. If yes, then does that mean JNDI somehow uses the DBCP internally?
When I have to create a DBCP for Swing applications, I had to first create an instance of GenericObjectPool and then create a connection factory object and finally a PoolableConnectionFactory object to create the Datasource which will be used to get a connection.

JNDI is a mechanism to pass objects from one part of the system to another (in technical terms across class loaders). This is most useful for classes and interfaces found in the Java Runtime like String or DataSource.
This means that in your case JNDI is just a transport mechanism and you need to have the actual connection pool defined elsewhere. Most web containers have a mechanism for defining a system wide connection pool, and JNDI then allows you to get to it.

Tomcat use a custom implementation of Apache DBCP and Apache Pool for the JNDI Datasource. These libraries are located in a single JAR at $CATALINA_HOME/lib/tomcat-dbcp.jar.
The main package is org.apache.tomcat.dbcp for avoid conflicts with the regular packages form Apache Commons.

JNDI is a mechanism of locating remote resources via lookup. It has nothing in common with connection pooling libraries. These libraries, of which c3p0, DBCP and BoneCP, are the most famous, allow you to create data sources with ability to pool connections and/or statements. If this data source is used within your application, you don't need to use JNDI, if it is located on a remote system (for instance, in Tomcat), you need to use JNDI to get access to the data source.
As a side note, why did you choose to work with old-school Servlet/JSP combo? It is a better idea to work with JSP successor, facelets that is a preferred view technology in JSF 2.x.
Another comment is to transfer management of your data source to a well-known framework. One direction might be to use an ORM, for example, Hibernate, to manage your data source (which was created with connection pooling in mind).

Initializing Memcached/JDBC Resources for a JAX-RS servlet

I have a service where I want to maintain data persistence in a mysql db using jdbc. While i have experience building jdbc apps and jax-rs apps in isolation, I've never combined the two. The question is, where does the build up and tear down required for the jdbc-type stuff go? Ordinarily i'd put the build up in a static block, or in a constructor, and id have have a cleanup method that gets called in a finally. this doesnt seem to work in the jax-rs framework- the constructor gets called at every invocation, and there is no place to my knowledge to put any clean up methods. unfortunately, there are sparse examples for combining the two technologies online, something i find surprising. Can you guys help me out?

As a general rule, to do things at the startup and shutdown of your web application, you should create custom ServletContextListeners and list them in your web.xml.
With JDBC resources in a WAR, often times you have your container (e.g. Tomcat, Websphere, etc.) create and manage a connection pool which can be shared with a number of web applications. You would define a resource-ref for a javax.sql.DataSource in your web.xml. Then there is a container specific method for defining and binding the JDBC DataSource to the resource-ref of your application.
I'm not familiar with Memchached and what is needed on startup/shutdown so this is only a guess. If you need to register/unregister with a Memcache server you might try having one or more env-entry tags defined in your web.xml which could be used by a custom ServletContextListener to do your bidding.

DI in an EJB/MDB Application

I'm currently developing a small EJB application running on IBM Websphere Application Server 7 (Java EE 5). The app mainly consists of one MDB listening for incoming MQ messages which are transformed and stored in a DB. Currently I'm using a lot of Singleton/Factories to share configurations, mappings, datasource lookups etc. But this actually leads to some very hard to test code. The solution might be using a (simple) DI framework like guice/spring to inject the different instances. The question is: Where to place the initialization/ setup code? Where is the main entry point of the application? How can I inject instances into the MDBs?

it might be worth looking at backing off from using Guice, and trying to work with the injection mechanisms already available with Java EE 5.
Regarding finding a suitable "startup point", unfortunately the EJB specification does not define a way where you can have a bean run at startup. However, the web profile of the EE spec does have one -- you can add a WAR to your application, and set a servlet listener component:
http://java.sun.com/javaee/5/docs/api/javax/servlet/ServletContextListener.html
You can set this to start whenever the application is loaded and started by the container (WebSphere). Watch out for classloader issues though.

Using Spring, you can do it via EJB3 interceptors, see http://static.springsource.org/spring/docs/3.0.x/spring-framework-reference/html/ejb.html#ejb-implementation-ejb3
Useful info on caveats are in the javadoc, make sure you read it: http://static.springsource.org/spring/docs/3.0.x/api/org/springframework/ejb/interceptor/SpringBeanAutowiringInterceptor.html

Any way to share session state between different applications in tomcat?

We want to split a working application in two different .war files in order to be able to update one app without affecting the other. Each webapp will have different a UI, different users and different deploy schedule.
The easiest path seems to be sharing the same session, so if app A set session.setAttribute("foo", "bar") app B will be able to see it.
Is there a way to share the HttpSession state for both apps in the same Tomcat instance?
Our app is running on a dedicated Tomcat 5.5, there are no other apps running on the same tomcat instance, so any security concerns regarding the session sharing are not a problem. We're running multiple Tomcat instances, but the balancer is using sticky sessions.
If it's not possible or this session sharing is a really bad idea please leave a comment.

You should not share HttpSession; but you can share other objects. For example, you can register an object via JNDI and access the same object in all your apps (databases use this to pool connections).

One thing to be aware of is that two web apps will use different classloaders. If you want to share objects, they need to use the same version of the class from the same classloader (or else you will get LinkageErrors). That means either putting them in a classloader shared by both web apps (system classpath for example) OR using serialization to effectively drain and reconstitute the object in the right classloader with the correct version of the class.

If you want to use Spring, there's a project called Spring Session:
https://github.com/spring-projects/spring-session
Quoting: "HttpSession - allows replacing the HttpSession in an application container (i.e. Tomcat) neutral way"

For Tomcat 8 I use the following configuration to share a session across 2 webapps:
conf/context.xml
<Context sessionCookiePath="/">
<Valve className="org.apache.catalina.valves.PersistentValve"/>
<Manager className="org.apache.catalina.session.PersistentManager">
<Store className="org.apache.catalina.session.FileStore" directory="${catalina.base}/temp/sessions"/>
</Manager>
...
</Context>
I deploy the same simple webapp twice log.war and log2.war:
/log
/log2
I can now log-in to /log and have the user displayed in /log2, this does not work with the tomcat default configuration.
The session value is set and read:
HttpSession session=request.getSession();
session.setAttribute("name",name);
HttpSession session=request.getSession(false);
String name=(String)session.getAttribute("name");
I used this project as example: https://www.javatpoint.com/servlet-http-session-login-and-logout-example
Most examples/solutions use a in-memory database which requires more setup work:
redis
hazelcast

If the two webapps are so closely coupled that they need to share objects, why are you splitting it in two? Even if you manage them somewhat independently any decent build management system should be able to create a single WAR file for deployment.
A solution like Aaron suggest with JNDI will work, but only if both webapps are running on the same server. If the units are tightly coupled and you are going to be running it on the same server anyway ... might as well have a single WAR.
If you really do want them to stand independently I'd seriously examine the data exchange between the two. Ideally you'd want them to only share relevant data with one another. This data could be passed back and forth via POST (or GET if more appropriate) parameters, you might even consider using cookies.

One way of doing this is described in this blog post: Session sharing in Apache Tomcat
Summary: Add emptySessionPath to the Connector configuration and crossContext to the Context

redison download
conf/context.xml
<Context sessionCookiePath="/">
...
<Manager className="org.redisson.tomcat.RedissonSessionManager"
configPath="${catalina.base}/conf/redisson.yaml"
readMode="REDIS" />
</Context>
conf/redisson.yaml
singleServerConfig:
address: "redis://<host>:6379"
sessionCookiePath="/" makes Tomcat use the same session id for different web apps.
RedissonSessionManager makes session to be persisted in 'shared space'
I was not able to achieve desired result with org.apache.catalina.session.FileStore PersistentManager in shared context.xml, I faced issues with session deserialization in background expiration monitor thread. It failed to deseriazile the session because it was using common classloader without webapp serializable models in classpath. Theoretically PersistentManager could be configured for each web app separately (to have proper classpath) in WEB-INF/context.xml but I failed to make it work.
org.apache.catalina.session.JDBCStore PersistentManage was promising because it expose last_access column for the session so it is not required to deserialize session_data, but it was saving app_name all the time causing same session id to be written as different rows for diffrent web apps. Thus session data was not stored in the shared place.
Spring Session has it`s own way to create session id. I was not able to find solution to force Spring Session to create same session id for different web apps.
Solution with core tomcat session id generation (with ability to generate the same for different web apps and RedissonSessionManager, which store data using session id as the only key and has it's own expiration mechanism) finally worked for me. The solution works perfectly with #SessionScope spring beans.

You can do by taking servlet context by your context root.
For retrieving variable.
request.getSession().getServletContext().getContext("/{applicationContextRoot}").getAttribute(variableName)
For setting variable:
request.getSession().getServletContext().getContext("/{applicationContextRoot}").setAttribute(variableName,variableValue)
Note: Both the applications should deployed in the same server.
Pls let me know if you find any problem

Tomcat 8 :
i had to do : <Context crossContext="true" sessionCookiePath="/"> in conf/context.xml
more details on config attributes here
and then to set the value(like #Qazi's answer):
ServletContext servletContext =request.getSession().getServletContext().getContext("contextPath")
servletContext.setAttribute(variableName,variableValue)
to get the value:
ServletContext servletContext =request.getSession().getServletContext().getContext("contextPath")
servletContext.getAttribute("user");

I developed session state server for tomcat using python.
Due to this I don't need to change the code already written for creating/accessing and destroying session. Also as there is separate server/service which is handling and storing session so not master cluster is needed. There is no session replication (as in tomcat clustering) in this case, rather this is session sharing between web farming.

You should not split your app that way in order by have high availability. You could deploy the whole app on many tomcat instances.

Best way to configure a Java enterprise application

I have a set of EJBs and other Java classes which need to be configured differently based on the system environment in which they are deployed: production, test, or lab. The configuration information includes stuff like URLs and database connection information.
We'd like to deploy the same exact product (EAR file) in each environment, and have the code then figure out where it is and what its configuration should be, without having to reach out to each deployment server in each environment to make changes.
What is the best way to configure all these components in a centralized, reliable, easy-to-maintain fashion?
Thanks for your thoughts.

The best, IMHO, is to use JNDI entries.
You may have to recode some parts of your application in order to use theses entries instead of plain vars, but with this setup:
Configuration is server-independant: each vendor provides its own implementation, but spec is a standard.
In a clustered environment, config can be persisted in a cluster-wide JNDI tree (see JBoss)
Configuration can be changed thru webadmin without restarting server.

How database connection pool information is stored / configured depends on the app server vendor. Put other variable stuff in property files on the classpath.

If you are deploying the exact same EAR to three different instances of a certain container than you will have to edit the deployment settings as there is no way that the deployment process could have any idea about which one of your three versions you would like to use at a particular deployment.
Deployment settings should go into JNDI entries as Piere-Yves said above.
If I were you, I would have my deployment-script (Ant?) properly populate the JNDI entries depending upon which environment you are deploying to.