[Design of JBoss Web Services] - Re: [Productivity] Level 0 - WS Culture
by maeste
Yesterday night I forgot two important thing I think you should include in docs. Probably they have to be distributed on all levels, but IMHO it's important to have them also at level 0.
1) a description of the importance of webservice in SOA. As said probably some IT manager will read your level 0 and it would understand why and where ws would be used. And you know SOA is a great product RH is developing...and a cool buz word ;).
2) An explaination of different stack implemented by jbossws. AFAIK you would provide at least native/metro/cxf. Users needs help to select the right one for their goals. Well it's probably mainly a level 1 docs, but some words in this level could be cool to explain to the users the JBoss added value leaving them free to select the best stack for their target.
Probably I forgot some others thoughts in this and upper level too..
View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4095886#4095886
Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4095886
18 years, 5 months
[Design of EJB 3.0] - Re: EJBTHREE-786
by ALRubinger
Keep coming around to this issue, and the collisions introduced by http://jira.jboss.com/jira/browse/EJBTHREE-1071 / EJB3 4.6.6 Bullet 1.
We can't allow a business interface to extend EJBObject/EJBLocalObject.
"bdecoste" wrote : I'm wondering what the trigger should be for us to support a 2.1 view (EJBObject)? Maybe just the existence of a Home interface?
I'm thinking that's it.
"wolfc" wrote : How about we take out the explicit binding of EJBObject in the proxy factories. If one of the interfaces extends EJBObject it'll be picked up again.
We've now taken the explicit binding of EJBObject out of the Proxy factories, but again, since the business interfaces cannot extend EJBObject, we're now in the position where no Proxy object ever implements EJBObject; local/remote interfaces are impossible.
Guess we'll have to bind EJBObject/EJBLocalObject back into the Proxies, and only if a Home interface is defined. This way, defining a Home interface completes a full EJB 2.1 view. Side effect is that the Proxy can now be casted to EJBObject - is this allowable?
Only other option is to put remote/local interfaces into their own Proxy in JNDI. We've talked about this before; seems excessive.
To this end. there's nothing in 4.7.1 that says that these classes cannot have overlap - ie. the class that implements the business interface may be the same class that implements the remote interface; in our case, the Proxy, which will do double-duty if Home is defined.
Thoughts?
S,
ALR
View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4095880#4095880
Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4095880
18 years, 5 months
[Design of JBoss Web Services] - Re: [Productivity] Level 3 - Transition / staging
by maeste
"alessio.soldano(a)jboss.com" wrote :
| * Speaking of tools, I'm thinking about something (let's say in the jmx-console) to enable/disable logging on every endpoint (see this post http://www.jboss.com/index.html?module=bb&op=viewtopic&t=115661).
|
Of course I agree ;P...well time it's a problem in this period, but I'll try to take a look soon as promised in that post.
anonymous wrote :
| Moreover we should cope with client side logging too, at least in the documentation; I'm thinking about explaining the need for tools like
| TCP-Mon
| http://ws.apache.org/commons/tcpmon/
| or Lms
| http://www.javalinuxlabs.org/lms/index.html
|
I'm convinced this kind of tool should be web based since, in my experience, a lot of time you need to trace communication with external customers and it's possible only with public IP address. Servers isn't used to have an X server active to use TCPMon....
Moreover LMS could be extended to communicate with the logging MBean described above and used to read (from a single console) server logs (or better all logs in the server farm) and client side logging if "proxy" feature is activated...
anonymous wrote :
| * Finally, speaking of QA tests, I would suggest users a web based tool for generic webservice invocation; developers might feel comfortable with java test cases or IDE tools, however those might be too difficult to be used by non-technical guys like QA testers. For example Wise (http://www.javalinuxlabs.org/wise/index.html) provides this functionality.
This sound like music to my ears...:)
As said in previous level it could be very useful also to generate unit test starting from QA tests, and in my experience (we use it internally since 3 years) it facilitate a lot the communication between QA and developers about ws bugs.
hoping it may help
View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4095816#4095816
Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4095816
18 years, 5 months
[Design of JBoss Web Services] - Re: [Productivity] Level 2 - Development
by maeste
"alessio.soldano(a)jboss.com" wrote :
| Speaking of samples, we think to provide meaningful advanced samples. These might cover the testcases I spoke about in the productivity introduction http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4095128#4094685, supporting tutorials in the documentation on advanced topics like security, attachments, addressing, etc.
|
Amazing...but please let me play with this meaningful advanced samples. What I'm trying to say is that this sample would be a ws sample, but to better understand what it does users have to play with the genarated application.
With "Play with" I mean they have to understand which services is provided (explaination in docs) call ws and see the results, read logs and understand what it's happening. In my experience too simple samples is useless as you said, but complex ones can become useless if users can't play with them.
IMHO Wise and LMS (you introduced in next level) can have a role in this "educational play", and SoapUI too of course. But the most amazing approach would be to develop dedicated web interface for this sample, giving to users opportunity to deploy and start to play, without hving to understand one more tool.
anonymous wrote :
| At the same time, we might focus on tools supporting the development phase of webservices. I'm thinking about a way to easily generate tests (JUnit or TestNG) for the users' services in a given project as far as an Eclipse project configuration with an Ant build file to run them all.
|
Last commit on wise (you were too busy to get it) went in this direction.
The idea is to generate test case using input provided by the user in the web interface. The assertion could use output received in web interface and validated by the user or not (leaving the developers free to write their own assertion). This approach could be very useful to generate regression tests starting from input/output of a QA reported bug.
I can explain a little more if you are interested in this idea.
Of course it's good for unit regression test, and not for massive or load tests, it would be annoying to insert billion of test case using a web interface :). Anyway I'm convinced some solution can be studied also for these situation using wise generated code just to have a test template and some standard or customizable parser to call this test template.
hoping it might help
View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4095813#4095813
Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4095813
18 years, 5 months
[Design of JBoss Web Services] - Re: [Productivity] Level 0 - WS Culture
by maeste
My two cents about level 0.
I think what you are saying would be useful explaining what webservices are, but probably most non-expert users need something clarifying what web service are not. I mean, simplifying a lot of course, something like "web service are not simple xml exchange, are not xml to parse with proprietary sw or read by human" and so on.
Another important thing to consider at this level is the target of the docs. Readers would be programmers, but probably also decision maker, IT manager, and maybe QA people who want to understand what they are using. Mainly IT manager (or better people who decide budget ;) ) is an interesting target to get attention and to drive to decide if ws is good for their project. It's also important at this level readers can get the idea and can decide IF ws is a good solution for their environment.
IOW you have to inform and evangelize, not to persuade to use ws technology. When readers would take the decision you would have to persuade them to use JBoss stack instead of any others of course, but it is a step forward :).
hoping it might help
View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4095797#4095797
Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4095797
18 years, 5 months
[Design of JBossCache] - Node removal not working with Invalidation?
by bstansberry＠jboss.com
I'm seeing weird behavior in my unit tests of Hibernate/JBC integration with invalidation, so thought I'd look for JBC testsuite tests of node removal with invalidation. Didn't find any. So, wrote the following which fails at the commented point.
Basically, a node removal doesn't seem to propagate. An equivalent test with replication instead of invalidation passed.
Am I missing something obvious here? If not I'll check in the test and raise a JIRA.
| public void testPessimisticNodeRemoval() throws Exception
| {
| nodeRemovalTest(false);
| }
|
| public void testOptimisticNodeRemoval() throws Exception
| {
| nodeRemovalTest(true);
| }
|
| @SuppressWarnings("unchecked")
| private void nodeRemovalTest(boolean optimistic) throws Exception
| {
| CacheImpl<Object, Object> cache1 = null;
| CacheImpl<Object, Object> cache2 = null;
| try
| {
| cache1 = createCache(optimistic);
| cache2 = createCache(optimistic);
|
| Node root1 = cache1.getRoot();
| Node root2 = cache2.getRoot();
|
| // this fqn is relative, but since it is from the root it may as well be absolute
| Fqn<String> fqn = Fqn.fromString("/test/fqn");
| cache1.getInvocationContext().getOptionOverrides().setCacheModeLocal(true);
| root1.addChild(fqn);
| assertTrue(root1.hasChild(fqn));
| cache2.getInvocationContext().getOptionOverrides().setCacheModeLocal(true);
| root2.addChild(fqn);
| assertTrue(root2.hasChild(fqn));
|
| assertEquals(true, cache1.removeNode(fqn));
| assertFalse(root1.hasChild(fqn));
|
| // THIS ASSERTION FAILS!
| assertFalse(root2.hasChild(fqn));
|
| assertEquals(false, cache1.removeNode(fqn));
|
| finally
| {
| // stop caches, etc.
| }
| }
View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4095794#4095794
Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4095794
18 years, 5 months