Category Archives: Code

How to set SVN global ignores

Leave a reply

There are some customizations that I make that pertains to my development computer only and because I work in a team environment I cannot commit in things that are customized towards my development environment. In order to do this, I do a global ignore on the file. Here’s how to do a global ignore. The subversion config file for Ubuntu 12.04 LTS or 12.10 is here:

sudo nano /etc/subversion/config

There is a section that looks like this

[miscellany]
### Set global-ignores to a set of whitespace-delimited globs
### which Subversion will ignore in its 'status' output, and
### while importing or adding files and directories.
### '*' matches leading dots, e.g. '*.rej' matches '.foo.rej'.
# global-ignores = *.o *.lo *.la *.al .libs *.so *.so.[0-9]* *.a *.pyc *.pyo
#   *.rej *~ #*# .#* .*.swp .DS_Store

You can uncomment that “global-ignores” part.

How to run multiple applications on a single tomcat

Leave a reply

I do this in Ubuntu 12.10 but I think it applies to Windows 7 configurations too. I wanted to run more than one application in a single instance of tomcat. I have to say this is not the best decision I have made.

You can load two apps in one server instance if you put them in different directories and have two different host names.

In my case, my two hostnames are app1.internal.kw.sg and app2.internal.kw.sg, yours will be different. You can edit your hosts here in your /etc/hosts

Both of my apps would be called app1 and app2. I put them into the corresponding directories, as specified in appBase. In my case, I placed them in:

  • /var/lib/tomcat7/app1
  • /var/lib/tomcat7/app2

You can open up your /etc/tomcat7/server.xml file. In Windows 7, it would be where ever your central tomcat configuration files are at. This is my server.xml file:

<?xml version='1.0' encoding='utf-8'?>
<Server port="8005" shutdown="SHUTDOWN">
	<Listener className="org.apache.catalina.core.JasperListener" />
	<Listener className="org.apache.catalina.core.JreMemoryLeakPreventionListener" />
	<Listener className="org.apache.catalina.mbeans.GlobalResourcesLifecycleListener" />
	<Listener className="org.apache.catalina.core.ThreadLocalLeakPreventionListener" />

	<GlobalNamingResources>
		<Resource name="UserDatabase" auth="Container"
			type="org.apache.catalina.UserDatabase"
			description="User database that can be updated and saved"
			factory="org.apache.catalina.users.MemoryUserDatabaseFactory"
			pathname="conf/tomcat-users.xml" />
	</GlobalNamingResources>

	<Service name="app1">
		<Connector port="8081" protocol="HTTP/1.1" connectionTimeout="20000" URIEncoding="UTF-8" redirectPort="8443" />
		<Engine name="app1" defaultHost="app1.internal.kw.sg">
			<Host name="app1.internal.kw.sg"  appBase="app1" unpackWARs="true" autoDeploy="true">
				<Valve className="org.apache.catalina.valves.AccessLogValve" directory="logs" prefix="localhost_access_log." suffix=".txt" pattern="%h %l %u %t &quot;%r&quot; %s %b" />
			</Host>
		</Engine>
	</Service>

	<Service name="app2">
		<Connector port="8082" protocol="HTTP/1.1" connectionTimeout="20000" URIEncoding="UTF-8" redirectPort="8443" />
		<Engine name="app2" defaultHost="app2.internal.kw.sg">
			<Host name="app2.internal.kw.sg"  appBase="app2" unpackWARs="true" autoDeploy="true">
				<Valve className="org.apache.catalina.valves.AccessLogValve" directory="logs" prefix="localhost_access_log." suffix=".txt" pattern="%h %l %u %t &quot;%r&quot; %s %b" />
			</Host>
		</Engine>
	</Service>
</Server>

You can refer to it. Unfortunately I cannot provide much support for this.

Disadvantages

Well, this method has a key disadvantage, each time you rebuild your Java classes, you need to restart the server and it would be better to be able to restart one application at a time rather than both applications together due to a longer loading time. Most of the time you will only make changes to one application anyway. It is better to have a configuration with two servers and each of them supporting just one application. That’s my take and I learnt it the hard way.

Too much magic in Java IDEs

Leave a reply

I am very bothered by the way a typical Java web site development works. Java website developments, through JSP (Java Server Pages) is largely supported by IDEs such as Eclipse and Netbeans. It is not in any way the easiest to utilize. And at worst these huge IDEs are too kludgy for my liking.

Take the current Java project that I am working on, it’s a website that has sign in and numerous data presentation tools. Everything .java compiles into a .class file and the remaining JSP files are deployed into this WAR file through Apache Maven into tomcat. With an IDE, everything seems seamless. You don’t even need to know how it works. Just by magic. You configure the servers and paths and it just works.

You could of course decide to dump WAR files again and again to the server paths and watch it automatically deploy after you do a manual restart of tomcat. Unfortunately that’s too unproductive, starting a tomcat server takes 2 seconds for me, and starting the application takes 10 seconds in debug mode. That’s unacceptably slow because it all adds up.

So, IDEs make you more productive by deploying the files in the server for you and compiles the .java files into .class and .jar files or something like that. Essentially you only need to restart the tomcat7 server if you updated the controller. Updates to the views do not require compilation. However — there’s always a caveat — you have the .jsp file edited on the server path rather than your workspace.

Now, under the hood of the IDE, each time you save a .jsp file, what you are essentially doing is:

  1. Saving the file in your workspace.
  2. Copying the file from your workspace to the server path.

Your workspace can be managed by revisioning tools such as SVN or git and the remnants don’t get copied over to the server space. It’s pretty well thought except that you really need a proper IDE set up to save you all the repetitive work.

So the reason why I am writing this is because I wanted to use Sublime Text and Sublime Text doesn’t have Java and tomcat integration features at all. At first this seemed surprising, after all Java is so common. Upon discovering what Eclipse and Netbeans actually does, I can’t help but to feel that there is a certain amount of over-engineering.

Forgive me if that’s not the way Java IDEs work. I’m uncovering new things every day still, so please correct my mistakes, I want to understand the platform better too.

Tomcat paths in Ubuntu 12.10

Leave a reply

And so I have the displeasure of working with Java — a platform I once swear to avoid for the rest of my life — and I had to have my Ubuntu 12.10 environment set up. This is also verified to be similar in Ubuntu 12.04.

If you install tomcat7 from the official repository using “sudo apt-get install tomcat7″, the following paths and commands are probably helpful to you too.

Your WAR files

WAR files (.war) typically go into:

/var/lib/tomcat7/webapps

Your log files

Your log files should be here:

/var/lib/tomcat7/logs

It should be called “catalina.out”. It’s useful to check here for troubleshoot Apache Tomcat. It doesn’t mean that tomcat’s successful start would mean the application to be working. catalina.out has saved me on several occasions.

Your tomcat7 configuration files

Server settings goes here:

/etc/tomcat7/server.xml

You might find the other properties files useful as well:

/etc/tomcat7

Your tomcat7 and catalina startup shell files

You might need to edit some .sh files like in my case to change Catalina options. You can find the shell scripts here:

/usr/share/tomcat7/bin

You can read more on setting up Catalina environment options.

Other notes

If you know of other paths worth including, do post a comment. While this is documented more for personal use, I hope you can find it useful too!

How to set Catalina options in Ubuntu 12.10

1 Reply

This is more of a personal note and a really specific topic. This tutorial assumes you are using Ubuntu 12.10 or Ubuntu 12.04 and installed tomcat7 package. If you haven’t already installed tomcat7, use:

sudo apt-get install tomcat7

After doing so, create and edit the file setenv.sh. Putting setenv.sh in CATALINA_BASE/bin allows you to keep your customizations separate.

sudo nano /usr/share/tomcat7/bin/setenv.sh

I set my environment variables (in my case) as in setenv.sh:

CATALINA_OPTS="$CATALINA_OPTS -Xmx1024m -XX:PermSize=256m -XX:MaxPermSize=512m"

If the file is present, catalina.sh will pick it up and add to the runtime parameters.

To restart tomcat7, use:

sudo service tomcat7 restart

Verifying that it works

After the server restart, run this to check:

ps aux|grep jar

This should return:

tomcat7  14140 63.8 11.4 4001076 918296 ?      Sl   12:26   1:37 /usr/lib/jvm/default-java/bin/java -Djava.util.logging.config.file=/var/lib/tomcat7/conf/logging.properties -Djava.awt.headless=true -Xmx128m -XX:+UseConcMarkSweepGC -Djava.util.logging.manager=org.apache.juli.ClassLoaderLogManager -Xmx1024m -XX:PermSize=256m -XX:MaxPermSize=512m -Djava.endorsed.dirs=/usr/share/tomcat7/endorsed -classpath /usr/share/tomcat7/bin/bootstrap.jar:/usr/share/tomcat7/bin/tomcat-juli.jar -Dcatalina.base=/var/lib/tomcat7 -Dcatalina.home=/usr/share/tomcat7 -Djava.io.tmpdir=/tmp/tomcat7-tomcat7-tmp org.apache.catalina.startup.Bootstrap start
kahwee   15564  0.0  0.0  13580   936 pts/1    S+   12:29   0:00 grep jar

You should be able spot your line of customization. And that’s the proper way to setting up Catalina options. You can also set JAVA_OPTS through this method too.

Degraded performance in Amazon EC2

Leave a reply

Amazon Elastic Compute Cloud (N. Virginia) or EC2 is down. This brought down Heroku and Reddit together with many others. It kinda serves as a reminder how reliant companies are on Amazon if anything.

What’s worse is that Amazon EC2 console is down too. Amazon says, “We are experiencing elevated error rates with the EC2 Management Console.”

The North Virginia facility seemed more unreliable than other availability sites. I used to run 2 EC2 instances in Singapore (Asia Pacific) for work and they’re much more reliable than this. There was a case where they had to do some hardware migration but that’s mostly minor.

Why does Rails do utf8=✓

Leave a reply

I noticed Rails apps always does utf8=✓ in their URLs. Rails at one point of time even placed a snowman unicode glyph. Here’s what Yehuda Katz has to say on this regard:

This parameter was added to forms in order to force Internet Explorer (5, 6, 7 and 8) to encode its parameters as unicode.

Specifically, this bug can be triggered if the user switches the browser’s encoding to Latin-1. To understand why a user would decide to do something seemingly so crazy, check out this google search: http://www.google.com/search?sourceid=chrome&ie=UTF-8&q=diamond+with+a+question+mark+in+it. Once the user has put the web-site into Latin-1 mode, if they use characters that can be understood as both Latin-1 and Unicode (for instance, é or ç, common in names), Internet Explorer will encode them in Latin-1.

This means that if a user searches for “Ché Guevara”, it will come through incorrectly on the server-side. In Ruby 1.9, this will result in an encoding error when the text inevitably makes its way into the regular expression engine. In Ruby 1.8, it will result in broken results for the user.

By creating a parameter that can only be understood by IE as a unicode character, we are forcing IE to look at the accept-charset attribute, which then tells it to encode all of the characters as UTF-8, even ones that can be encoded in Latin-1.

Keep in mind that in Ruby 1.8, it is extremely trivial to get Latin-1 data into your UTF-8 database (since NOTHING in the entire stack checks that the bytes that the user sent at any point are valid UTF-8 characters). As a result, it’s extremely common for Ruby applications (and PHP applications, etc. etc.) to exhibit this user-facing bug, and therefore extremely common for users to try to change the encoding as a palliative measure.

All that said, when I wrote this patch, I didn’t realize that the name of the parameter would ever appear in a user-facing place (it does with forms that use the GET action, such as search forms). Since it does, we will rename this parameter to _e, and use a more innocuous-looking unicode character.

Very funky although this has since become my standard way of determine if the application is running on Ruby on Rails.