Win a copy of Testing JavaScript Applications this week in the HTML Pages with CSS and JavaScript forum!

Nitin Mehhta

Ranch Hand
+ Follow
since May 18, 2004
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 Nitin Mehhta

Just to add to my question, I know such a dynamic validation is possible in DOM3 but the environment I am using is DOM2 jdk1.4_2
14 years ago
I'm facing a problem for which I am unable to find a correct solution. I have a XML file which I read into memory and make a DOM (using validating parser). Now I start removing certain nodes from the DOM. I want to check that each Node removal is not causing the document to be invalid against the DTD. I want to check this as and when i'm removing the nodes and if the document is becoming invalid because of theremoval then I do not want to remove the node. The only way I can think of achieving that right now is removing each node then writing the file back and again validating it by parsing through a validating parser. But that will hit the performance a lot. Can anyone suggest a better alternate, I'm sure they would be a better way.

Thanks in advance.
Nitin
14 years ago
Thanks Chris for elaborate response to my query. I really appreciate your taking time out for this. Let me correct my self when I said EJB vs Hibernate. I actualy ment EJB Entity Beans vs Hibernate. Sorry for the confusion caused. As far as my analysis over there is concerned I read in one of the articles (http://www.roseindia.net/hibernate/hibernate_architecture.shtml) that Hibernate comes with its own Connection Pooling module and transaction Manager module which does not have the desired performance. Your assumption of 50 TPS holds true for our application. What I did not know was that this figure does not qualify as high TPS. That would certainly change things a bit. I was not aware of the IBatis project and would certainly have a look there. Just to summarize from the discussion, my concerns were mainly about a) connection pooling - for which you have advised to make use of the App server,(instead of hibernate connection pooling module or any other) b) transaction - which in this case can be managed by JDBC itself and c) performance - which again is not a major factor with the given TPS value. Once again thanks for the insight.
Thanks Stan, Chris and Shan for the valuable comments. We also explored the idea of POJO first but then we thought of utilizing EJB out of the box facility. I would go back and look again at the real worth vs weight of EJB. As far as Hibernate is considered we did a EJB vs Hibernate analysis and found out that performance for connection pooling and transaction management using Hibernate was not up to mark. Moreover in our design ORM was not as valuable as transaction management (so we chose Session EJB). I would seriously give a solid cosideration to Spring now. Again thanks for the review and suggestions.
Hi,

I�m developing a system which will facilitate in conducting online tests for the various software streams. The major challenge while designing this application is of a large number (say 200) of concurrent users. This would be a web application built on J2ee framework. I�m using struts framework which will take care of the viewer and the controller aspects of the application. As there will be a high transaction rate (due to various simultaneous tests) I have chosen EJBs (Stateless session beans) to get the benefit of the application server transaction handler. My Action Servlet will call an Action class which will delegate the work to a facade class. The facade class will serve as the EJB client and call the EJB business methods. I have made a database helper class which has methods to facilitate various database interactions like selects, inserts, updates etc. Inside the EJB business method I�m calling methods from this database helper class to do the database interaction. I�m passing the result back to the web server (fa�ade class) via Value Objects. I�m at the architecture design phase for the application. As this is my first venture in developing an EJB compliant web application please let me know if the architecture I have followed is correct or flawed. Also I would really welcome any suggestions\improvements in the design.

Thanks in Advance.
This post is to thank everyone participating in this very good and helpful forum. I took my SCBCD exam today and got 64/70 (91%). I found the exam to be very easy in comparision to the test simulators ( i used EJBPlus ). I am mentioning this because like many people I had this phobia that the real thing would be more difficult than the simulation tests, and so I kept on procrastinating my exam. ;-) Once again a big thanks to all of you and best of luck to the people who are thinking of taking this exam.

Nitin
SCJP1.2
SCWCD1.3
SCBCD1.3
15 years ago
Hi All,
I have a java webservice, running on websphere4.0, which runs out of memory very frequently. Please advice as to what I can do to rectify the situation. I'm very new to webservice. If this is not the correct forum for this question then please move it to some place relevant.

Thanks
Nitin
I think both are correct. Member Of is also valid in EJB-QL...
it tests if a single valued entry is there in a collection-valued path expression.
Thanks for posting this question and replying to it as I had attempted the same question in one of the simulation exams where I had also marked the same answer given By Giju. But my answer was marked wrong. Just to add my 2 cents to this ... I think I read it somewhere that whenever the method is not returning the exact component interface we need to use the narrow method.
And yes we should not throw Runtime exceptions...it is just that something which we are not expecting happens then these exceptions are generated from the running code. So they are propogated automatically. I think by that the specs means that we should not explicitly throw any runtime exception.
Actually according to my interpretation of what I have read I think in the cases where the delivery system does not get a acknowledgement of the message a redelivery takes place. For a CMT the acknoledgement is always tied up with the transaction. So if transaction rolls back we get the redelivery...even if it is done by calling setRollBackOnly()... where as in the case of a BMT Message bean the transaction is tied to the method completion. So if we setRollBack using code and the method returns fine...i mean completes then no redelivery happens. Only if we get a system exception ..where the method fails to complete we get a redelivery.
Hi,
I have two questions.
1) Can a method be a part of the exclude-list as well as be present in one of the method-permission tags. More relavantly..what will happen if we say
<exclude-list><method>*</exclude-list></method> and then we specificly assign a method to a role using a method-permission tag....what will be the output.

2) I read somewhere that if a CMT stateful session bean implements SessionSynchronizer interface then the only allowed trans-attributes are "Mandatory" "Required" and "RequiresNew". Is this true. I know it makes sense to have a transaction in such a case but...is this restriction true....purely from the point of view of the exams. :-)

Thanks in advance
Adding to this confusion the same book at page 550 says that a Stateless Session Bean can never have RemoveException. Which I feel though in the context of the discussion (in the book) is relevant, but still it is incorrect. At that time probably the author wants to communicate that we cannot get a RemoveException from the component interface of the Stateless Session Bean. Please do correct me if my assumption is wrong.
yes absolutely. I think you had posted a similar query sometime back. It is some how confusing and the array of wrong answers given add to the woes. For most of the people using the HF book a further reiteratement could be found at page 130 and page 559 -> second scenario.
I mean to say that if you have explicitly mentioned in your ejb-ref that you would be using a remote/local reference then you should do accordingly. But say you have not defined the ejb-ref tag in your DD then you have to give the actual JNDI name in your code. And in that case if your entity bean and session bean are on the same jvm then you don't need to do narrowing. The basic idea behind my answer is that ejb-ref tag is not a necessity. If you have it then you should abide by it.