Hi, all
I have been preparing for
SCEA Phase I and will take the exam in this week.
I have some question listed as follows and along with answers provided by the source who created the questions. For some questions, I have different opinions to the answers, and for some questions I just don't know the reason behind the answers. Could someone please shed some lights on them? Many thanks!
QUESTION NO: 1
The exhibit shows the communication between four EJBs.
Customer(Session
EJB) ->call findByPrimaryKey() on -> Account (Entity Bean) -> call findByPrimaryKey() on -> Order(Entity Bean) -> call findByPrimaryKey() on -> ShoppingCard (Entity Bean).
Which two changes to the existing architecture will increase performance? (Choose two)
A. Convert all entity beans to stateless session beans.
B. Use bean-managed persistence for each entity bean.
C. Have Customer call Account, Order, and ShoppingCart directly.
D. Have each EJB hold a reference to the EJB it calls, instead of calling findByPrimaryKey().
Answer: C, D
Can any one let me know why option C can improve the performance? QUESTION NO: 2
The current architecture has
servlets making calls to EJB. The system has grown to over 200 servlets causing development and configuration problems. Because the system is still evolving, the APIs to the EJB have changed quite often. This has caused the servlet writers to adapt the new APIs every time a change is made.
Which two statements about this problem are true? (Choose two)
A. Introducing a configuration control system will fix the problem.
B. This is an example of brittle code dependencies and tight coupling.
C. This problem can be corrected by converting all entity EJB to stateless session EJB.
D. Introducing a client-side abstraction which sits between the servlet and FJB will break the tight coupling.
E. Having the servlets communicate with EJB using JMS will prevent changing servlet code every time an EJB API changes.
F. Introducing a session bean, in front of all EJB, which mimics the EJB APIs will prevent changing servlet code every time an EJB API changes.
Answer: B, D
Can any one shed a light why option E is not correct? Isn't messaging the one of the approaches to promote loose coupling? Of course, we need MDB at backend. QUESTION NO:3
xxx.com, a shipping company, is building an enterprise system to track the location of packages. One part of the tracking system is a network of wireless inventory devices. The devices can only be accessed using a custom, synchronous TCP/IP protocol. How should you encapsulate interaction with the wireless inventory system?
A. With an EJB session bean that uses a JMS to interact with the inventory system.
B. With a
Java class that uses custom Java networking to interact with the inventory system.
C. With a distributed CORBA object that uses IIOP to interact directly with the inventory system.
D. With an EJB entity bean that uses container-managed persistence to encapsulate the inventory system.
Answer: C
Why option B is not correct. This question does not say anywhere CORBA will be supported by the wireless system. QUESTION NO: 4
A steel industry association has hired you, a network consultant, to create a B2B architecture for the steel industry. What does this mean?
A. You have to create a portal to allow consumers to order steel products.
B. You have to create a portal to
exchange news and information about the steel industry.
C. You have to create an infrastructure to allow the exchange of product catalog information between steel companies.
D. You have to create an architecture to allow the exchange of order information between steel companies order systems.
Answer: D
Would anyone please explain why option c is not correct? Anyway, I didn't find a precise definition of B2B in any SCEA books/notes. QUESTION NO: 5
Which three are aspects of an application are most likely to be determined at runtime based on the user�s declared nationality of locale? (Choose three)
A. Currency formats.
B. Network protocols.
C. Textual output messages.
D. Calculations and/or algorithms.
E. Server host names and/or addresses
Answer: A, C, E
Why E is correct? Anyone think D is a better choice? QUESTION NO: 6
XXX Grapes, a vineyard, is developing an online ordering system to sell wine online. Because of local sales restrictions, the vineyard is working with local wine stored to deliver wine to the customers. The requirements for the system are:
� Customers keep their choices in a �wine list�.
� Requests are fulfilled as the vintage become available.
� Local wine shops track wines that have been ordered by customers in their area.
� All order and product data is stored in an RDBMS.
� Both customers and local wine shops use the Internet to access the system.
� The system must be scalable and secure.
Which two interactions must be secure? (Choose two)
A. Customers adding wine to their wine list.
B. The application server generating a sales report.
C. The application server updating the order information into the DB.
D. The local wine stop checking what wines have been ordered locally.
Answer: A, D
I saw this question has been discussed in other posts. Here just provide the official answer for reference and discusson. QUESTION NO: 7
Which statement describes a normal default security restriction applied to classes loaded from untrusted
sourced by Java-enabled browsers?
A. Untrusted classes cannot load trusted classes.
B. Untrusted classes cannot create server sockets.
C. Untrusted classes cannot load untrusted classes.
D. Untrusted classes cannot initiate any network connections.
E. Untrusted classes cannot display windows outside of the browser.
Answer: B
Could anyone explain why option C and E is not correct?