Ed Napier

Greenhorn
+ Follow
since Mar 10, 2010
Ed likes ...
Oracle Java
Merit badge: grant badges
Biography
I started developing with Java in 2004, and have continued largely self-taught or partnered.  I use IBM RAD at work and (occasionally) Eclipse at home.
For More
Cincinnati, OH
Cows and Likes
Cows
Total received
0
In last 30 days
0
Total given
0
Likes
Total received
0
Received in last 30 days
0
Total given
0
Given in last 30 days
0
Forums and Threads
Scavenger Hunt
expand Ranch Hand Scavenger Hunt
expand Greenhorn Scavenger Hunt

Recent posts by Ed Napier

After years of using RAD and WebSphere, I just started using Eclipse Luna, Gradle 2.21 and WebSphere Liberty Profile 8.5 (as directed by my boss, so this tool set is a given). The problem I am running into is the Liberty server is not refreshing with the latest build of my controller. It seems to be caching the older version, even after a stop and start.

For example, I replaced one class constructor with one that takes a different argument, and then updated the call to the constructor in the controller. (This was with the server stopped.) I then did a gradle clean build, which completed without issue. Finally I set the server to clean on next start, then started the server in debug mode. When I started the app, I got an error that the controller couldn't find the old constructor.

Unfortunately google's not been much help in this case (as it sometime's isn't). Any ideas for places to look would be most appreciated.

Thanks, Ed
9 years ago
The initial client project is Java based, and future clients will most likely be Java based also.

I do already have a wsdl, so I will try your suggestion and will post the result.
9 years ago
I made a webservice with RAD 8, and it seems to work fine. I've been able to load it and test it using the on-board Web Services Explorer, and all looks hunky-dory.

So next I copied the generated WSDL to my application that will ultimately use this service. When I tried to create the client from the WSDL, I get this at the top of the dialog window:
"The WebSphere Application Server v8.0 server does not support the client project projectName."

I'm guessing it's a version problem in the client project, but where do I look? Google's been no help nor have the Red Books.
9 years ago
Thanks all.
10 years ago
In most cases they are classes that have one or more elements reading and/or writing to an Oracle db via a JNDI alias to a JDBC connection. A handful connect to some web service and the remainder are Struts 1.3 actions.
10 years ago
Over the years I've written boatloads of Junit method tests but never got around to actually doing anything in container (properly). For a variety of reasons, loading up cactus and doing in-container testing just seemed harder to work out than just running with breakpoints and logging things in a spreadsheet. However, as I've gotten busier, and processes and procedures have become more involved and time-consuming, that bias on my part isn't an option any more.
Unfortunately, Cactus is now pretty much dead. I couldn't even find a reasonably recent download of it any more.

What suggestions do people have for an easy to set up in-container test bed that works well with IBM RAD 8?
10 years ago
Thanks, Paul. I tried just commenting out the log4j config lines in our properties db and the primary apps retained control over the logger. The possible downside is that it didn't appear the help apps were logging anything into the primary logger stream, but I'm not sure that there were any messages to log. I'll run an experiment case in a couple of days to clean up any loose ends.
Making a long story short, while trying to debug an an issue in one of my applications, I noticed my debug statements were not writing to my logfile. In addition, the logfile would only be updated after I had just restarted my development server (WAS).

No errors were showing up anywhere, so I wrote a little wrapper around my method that wrote logger info to the console. That was when I noticed the configuration values at the end of the method had been replaced. Ultimately I was able to narrow the before and after until I had shown that a call out to class in a helper application (on the same server) had overwritten the previous values, including the output file. Sure enough, I went to that other file and there was the rest of the logger debug info from my application, as well as the mysterious "subsequent run" logs.

I have since found this condition in at least one other application using a different helper app. Since I am but one of about a dozen programmers, I assume there may be more.

Each application has its own log4j configuration and log file, including most of the helper apps. Most of those have a rootLogger setting. Nearly all are running inside the same JVM.

Googling didn't really help (at least the way I'd formed the query), so I'm looking for either things to try or places to look.