Resin as a PaaS: Amazon RDS/EC2 Web App Deployment with MySQL JDBC database

From Resin 4.0 Wiki

Revision as of 00:00, 6 April 2012 by Rick (Talk | contribs)
Jump to: navigation, search

[IN PROGRESS BEING UPGRADED to 4.0.28]

This is a continuance of our earlier tutorial on Resin cloud deploy on Amazon EC2. The first part of this tutorial could be used to see how to configure Resin for production deployments. We create a simple web application that uses MySQL.

The first part of this tutorial just shows how to use a database from JNDI. It shows how to install the JDBC jar files and how to configure Resin datasources. There is nothing cloud centric per se about the first part. It also shows how to do remote configuration of Resin servers.

Then we show you how to setup Amazon RDS for MySQL. Then configure Resin to talk to the RDS MySQL instance.

These tutorials are related:

Part 1 Resin_Cloud_deployment_with_Amazon

Part 2 Resin_Cloud_deployment_with_Amazon_WS_EC2_and_Euca_Tools


Contents

Tutorial Status

This tutorial is in progress. It is being updated from 4.0.24/4.0.26 to 4.0.27 and 4.0.28. Improvements were made to Resin to support remote cloud deployment and configuration.

We assume your development box is Ubuntu or Debian and your Amazon AMI is Amazon Linux (CentOS 4.5 more or less/Red Hat Enterprise).

Install MySQL on local box

First let's install MySQL create a user called bloguser with a password. Then let's create a database called blogdb.

$ sudo apt-get install mysql-server

Startup MySQL client tool to create a new user and database:

$ mysql -u root -p

Next create a user called bloguser with the pass roofoo.

mysql> GRANT ALL PRIVILEGES ON *.* TO 'bloguser'@'%' IDENTIFIED BY 'roofoo' WITH GRANT OPTION;

Output:
Query OK, 0 rows affected (0.00 sec)

Next create a database called blogdb.

mysql> create database blogdb;

Output:
Query OK, 1 row affected (0.00 sec)

Use blogdb database and create a table called blog.


mysql> use blogdb;

Output:
Database changed

Create a table called blog.

mysql> CREATE TABLE blog (id INT NOT NULL PRIMARY KEY AUTO_INCREMENT, 
                                            title VARCHAR(80), message VARCHAR(2000));

Output:
Query OK, 0 rows affected (0.01 sec)

mysql> exit;

Output:
Bye


Use Roo to switch to MySQL

In the examples folder start up roo and switch to using MySQL.


$ roo

roo> jpa setup --database MYSQL --provider HIBERNATE --databaseName blogdb

Output:
Updated SPRING_CONFIG_ROOT/database.properties
Please update your database details in src/main/resources/META-INF/spring/database.properties.
Updated SPRING_CONFIG_ROOT/applicationContext.xml
Updated ROOT/src/main/resources/META-INF/persistence.xml
Updated ROOT/pom.xml [added dependency mysql:mysql-connector-java:5.1.18; removed dependency com.h2database:h2:1.3.161]

Find and modify the database properties for this application. (Later we will set this up in JNDI of Resin, but for now just modify the properties file).



$ find . -name "database.properties"

Output
./src/main/resources/META-INF/spring/database.properties

$ emacs ./src/main/resources/META-INF/spring/database.properties

Add the user name and password that we setup earlier.

#Updated at Thu Dec 01 12:57:30 PST 2011
#Thu Dec 01 12:57:30 PST 2011
database.password=roofoo
database.url=jdbc\:mysql\://localhost\:3306/blogdb
database.username=bloguser
database.driverClassName=com.mysql.jdbc.Driver

Deploy and test what we have done so far

Now let's test out what we have so far. Build the project and deploy it to resin.

$ mvn package


Output:
...
[INFO] BUILD SUCCESS

$ resinctl deploy --name blogm ./target/blog-0.1.0.BUILD-SNAPSHOT.war 

Output
Deployed production/webapp/default/blogm from ./target/blog-0.1.0.BUILD-SNAPSHOT.war to hmux://127.0.0.1:6800

Since this is a local development box, you could just copy the war file over.

$ sudo cp ./target/blog-0.1.0.BUILD-SNAPSHOT.war /var/www/webapps/

Now startup the blogm app and add a blog entry.

$ firefox http://localhost:8080/blogm/

If you like, you can fire up the mysql client and see if your new entry and blog table is in MySQL.

$ mysql -u root -p
Enter password: 

mysql> use blogdb;

mysql> show tables;
+------------------+
| Tables_in_blogdb |
+------------------+
| blog             |
+------------------+
1 row in set (0.00 sec)

mysql> select * from blog;
+----+---------+-------+---------+
| id | message | title | version |
+----+---------+-------+---------+
|  1 | sdf     | sadf  |       0 |
+----+---------+-------+---------+
1 row in set (0.00 sec)

mysql> exit

Now change Resin and Roo to use JNDI instead of Spring configured datasource

Resin can monitor and manage database connection better than Spring alone. In order to do this you need to JNDI. It also good to enable JNDI for production so that a select few in operations can control the user name and password of the production system databases. Having JNDI allows your dev box and your prod box to have different usernames and passwords as well without rebuilding a special war file for each.

In this tutorial our dev box is Ubuntu and our prod box is Amazon AMI in the sky.

First tell Roo to use JNDI as follows:

$ roo
    ____  ____  ____  
   / __ \/ __ \/ __ \ 
  / /_/ / / / / / / / 
 / _, _/ /_/ / /_/ /  
/_/ |_|\____/\____/    1.2.0.RC1 [rev dcaa483]

...

roo> jpa setup --database MYSQL --provider HIBERNATE --databaseName blogdb --jndiDataSource jdbc/blogdb
Output:
Updated SPRING_CONFIG_ROOT/applicationContext.xml

roo> exit

Notice the addition of --jndiDataSource jdbc/blogdb.

Let's see the change it made.


$ find . -name applicationContext.xml
./src/main/resources/META-INF/spring/applicationContext.xml

$ grep -A 5 -B 5 jdbc ./src/main/resources/META-INF/spring/applicationContext.xml 
    </context:component-scan>
    <bean class="org.springframework.orm.jpa.JpaTransactionManager" id="transactionManager">
        <property name="entityManagerFactory" ref="entityManagerFactory"/>
    </bean>
    <tx:annotation-driven mode="aspectj" transaction-manager="transactionManager"/>
    <jee:jndi-lookup id="dataSource" jndi-name="jdbc/blogdb"/>
    <bean class="org.springframework.orm.jpa.LocalContainerEntityManagerFactoryBean" id="entityManagerFactory">
        <property name="persistenceUnitName" value="persistenceUnit"/>
        <property name="dataSource" ref="dataSource"/>
    </bean>
</beans>

Notice the line that has < jee:jndi-lookup id="dataSource" jndi-name="jdbc/blogdb".

Now we need to add a JNDI entry for our MySQL database to Resin and deploy the JDBC driver to Resin.

$ mkdir database-conf
$ cd database-conf


Create the file database-conf.xml


And add the contents as follows:


<resin xmlns="http://caucho.com/ns/resin"
      xmlns:resin="urn:java:com.caucho.resin">

<database jndi-name="jdbc/blogdb">
        <driver type="com.mysql.jdbc.Driver">
                 <url>jdbc:mysql://localhost:3306/blogdb</url>
                 <user>bloguser</user>
                 <password>roofoo</password>
         </driver>
</database>

</resin>


Find and copy the mysql jar file from the maven local repo to Resin as follows:

Find it:

$ find ~/.m2/repository/ -name "*mysql*.jar"
/home/rick/.m2/repository/mysql/mysql-connector-java/5.1.18/mysql-connector-java-5.1.18.jar


Now copy it to database-conf dir.

  $ sudo cp ~/.m2/repository/mysql/mysql-connector-java/5.1.18/mysql-connector-java-5.1.18.jar .

When you are done, you should have something like this:



$ pwd
/home/rick/examples/blog/database-conf

$ ls
database-conf.xml  mysql-connector-java-5.1.18.jar

$ cat database-conf.xml 
<resin xmlns="http://caucho.com/ns/resin"
      xmlns:resin="urn:java:com.caucho.resin">

<database jndi-name="jdbc/blogdb">
        <driver type="com.mysql.jdbc.Driver">
                 <url>jdbc:mysql://localhost:3306/blogdb</url>
                 <user>bloguser</user>
                 <password>roofoo</password>
         </driver>
</database>

</resin>

Now we can deploy this whole directory, jar file and configuration file to Resin in one swope using resinctl config-deploy. First let's use resinctl help system to give us some more information.


$ resinctl help config-deploy
usage: resinctl config-deploy [--options]

  deploys a configuration directory or jar file

where options include:
  --address <ip>         : IP address or host name of (triad) server
  --conf <file>          : alternate resin.xml file
  --log-directory <file> : alternate log directory
  --mode <string>        : select .resin properties mode
  --password <password>  : admin password for authentication
  --port <port>          : IP port of (triad) server
  --resin-home <dir>     : alternate resin home
  --root-directory <file> : alternate root directory
  --server <id>          : id of a server in the config file
  --user <user>          : admin user name for authentication
  --user-properties <file> : select an alternate $HOME/.resin file
  --verbose              : produce verbose output


If we don't specify the address and the port, then we should deploy locally to our dev box. Later when we want to deploy to Amazon WS, we just need to specify its address and port of our remote server in the cloud.

First let's deploy our config to our dev box.



$ cd ..

$ ls

database-conf  log.roo  pom.xml  src  target

$ resinctl config-deploy database-conf/

Deployed production/config/resin from database-conf/ to hmux://127.0.0.1:6800


To verify that our deploy worked, lets run resinctl config-ls, and resinctl config-cat:



$ resinctl config-ls

database-conf.xml
mysql-connector-java-5.1.18.jar

$ resinctl config-cat database-conf.xml

<resin xmlns="http://caucho.com/ns/resin"
      xmlns:resin="urn:java:com.caucho.resin">

<database jndi-name="jdbc/blogdb">
        <driver type="com.mysql.jdbc.Driver">
                 <url>jdbc:mysql://localhost:3306/blogdb</url>
                 <user>bloguser</user>
                 <password>roofoo</password>
         </driver>
</database>

</resin>


Deploy and test what we have done so far

Now let's test out what we have so far. Build the project and deploy it to resin.

$ mvn package


Output:
...
[INFO] BUILD SUCCESS

$ resinctl deploy --name blogjndi   ./target/blog-0.1.0.BUILD-SNAPSHOT.war 

Output
Deployed production/webapp/default/blogjndi from ./target/blog-0.1.0.BUILD-SNAPSHOT.war to hmux://127.0.0.1:6800

Now startup the blogm app and add a blog entry.

$ firefox http://localhost:8080/blogjndi/

Poke around with mysql client tool and see if your new blog entry is there.

LEFT OFF HERE.

Setting up variables

I don't like needing to modify XML all of the time. You can setup EL variables in database.xml that allow you to set key name value pairs in the resin.properties file as follows:

/etc/resin/resin.properties Add three properties to resin.properties

blogdb.url : localhost
blogdb.user : bloguser
blogdb.password : roofoo


/etc/resin/local.d/database.xml Change database.xml to use new properties.

<resin xmlns="http://caucho.com/ns/resin"
      xmlns:resin="urn:java:com.caucho.resin">

<database jndi-name="jdbc/blogdb">
        <driver type="com.mysql.jdbc.Driver">
                 <url>jdbc:mysql://${blogdb.url}:3306/blogdb</url>
                 <user>${blogdb.user}</user>
                 <password>${blogdb.password}</password>
         </driver>
</database>
</resin>

This will be nice because later we can set the user-data for the VM and point to the Amazon RDS database.

Setup RDS database

Setup an RDS database using MySQL 5.1.7 (or as close to your MySQL as possible).

I use the same username password to make it simple. The only part that is going to change for sure is the URL.

Using the AWS web console modify the user-data of the resinserver we deployed earlier and add three properies.

user-data on resin server


blogdb.url : blogdb.cvolnlau763z.us-east-1.rds.amazonaws.com
blogdb.user : bloguser
blogdb.password : roofoo

Note that the URL should be the url of the Amazon RDS database that you just setup. Also you need to setup a database security group that allows the group of your vm instance access the database.


Start up the VM instance that we installed Resin on earlier and create a database.xml file in the /etc/resin/local.d/ as follows:


/etc/resin/local.d/database.xml Create database.properties on remote Amazon instance.

<resin xmlns="http://caucho.com/ns/resin"
      xmlns:resin="urn:java:com.caucho.resin">

<database jndi-name="jdbc/blogdb">
        <driver type="com.mysql.jdbc.Driver">
                 <url>jdbc:mysql://${blogdb.url}:3306/blogdb</url>
                 <user>${blogdb.user}</user>
                 <password>${blogdb.password}</password>
         </driver>
</database>
</resin>


Next install the mysql driver on the Amazon instance.

$ sudo yum install mysql-connector-java
$ sudo cp /usr/share/java/mysql-connector-java.jar /usr/local/share/resin/lib/

Now deploy the war file to the Amazon AMI

 $ resinctl deploy target/blog-0.1.0.BUILD-SNAPSHOT.war -name blogdb -address 107.22.13.146 \
                                                                     -port 8080 -user admin -password roofoo

Now test it out: http://ec2-107-22-13-146.compute-1.amazonaws.com:8080/blogdb/blogs/1

It worked for me. :)

Personal tools
TOOLBOX
LANGUAGES