1。查一下API和众多的例子吧。
2。同1。也是API的问题。
3。天啊，这个问题太无耻了，你就画个图，把每个可能的情况都说说吧。
4。第一问和第二问实际就是要你吹吹Application Server的好处嘛，就那几个要点嘛，应该本本EJB的书都有，
还有，你说Tomcat是Application Server，千万不要给面试你的人听到，他会马上拿来候选名单，然后在你的名字上划个大红叉叉滴，：)

Trying the question, the algorithm to build a stack using linked list is at the end. hope it could help

a linked list has:

1. properties:
a. Node start
b. Node prev
c. Node current
2. methods:
a. add_before_current(object x)
b. current()
c. remove()
.....
for example
1. building the linked list
add_before_current("1")
add_before_current("2")
add_before_current("3")
//the linked list should be 3->2->1
//the current is pointing to "3"
2. obtaining the current node
current()
3. removing the current node
remove() //obtain "3"
remove() //obtain "2"
remove() //obtain "1"
//end

remark: a stack is a data structure which is "First In Last Out"

implementing the stack:
1. properties:
a. Node topofStack
.....
2. methods:
a. push(object x){
add_before_current(x);
}
b. pop(
current();
topofStack=remove();
)

In general: both are data structures

a few differences are:
1. array has to be fix size, however, binary tree can grow as many as possible
2. array is easy to implement, but complex in binary tree
3. regarding operations (insert, remove, findMin, findMax, search....) in an array and binary tree, usually, the worst-case complexity of the first one is n, but could be log(n) in a binary tree, unless a binary tree is a linked list

third point is the key between the comparision

hope it helps

one of problems may be: the client machine is abnormal caused by OS exception, viruses, low-speed internet connection and so on.

I won't say it is servers' problem because only one client has a degraded reponse time. as a developer or administrator who will believe servers have a problem only if more evidences in a problem are collected.

本文发表在 rolia.net 枫下论坛1)stack extend linkedlist {
private list = new LinkedList();
public void push(object o) {
list.add(o);
}

public object pop() {
return list.removeLast();
}
}
If that was to ask memory management concept in general, that would be too many topic to cover.
Such as the difference between implementing stack in linkedlist and array in terms of memory management. Seting heap size by
xms/xmx/perm/maxperm, different type of jvm (classic/server/hotspot), use profiling tools to diagnose memory leak and/or inefficient memory usage,
memory usage tricks, garbage collection timing, use of finalize method etc. That's too huge topic.

2)not exactly sure what do you mean by compare a array, what about b-tree? compare what to what?
3)cluster and load balancing http://e-docs.bea.com/wls/docs70/cluster/overview.html#1003768
4)this is not a yes-no questions. I think they are just trying to see how much you know about all sorts of servers. Some popular
ones are oc4j, websphere, weblogic. For exmaple, websphere is powerful to handle different version of classes use its confusing
and powerful classloader and it has its own jvm. WEblogic use sun's jvm and doesn't have much choices for class loader. OC4j
has some different settings for SOAP descriptor comparing with other servers. WEblogic is buggy in older version, it doesn't even
fully support J2EE spec. Size of jsp page matters in some server. WEblogic use thread pooling which might be an issue for some apps in terms of memory management blah,
blah, blah... It's all about how much practical experience you have. I am assuming you are looking for a development role.更多精彩文章及讨论，请光临枫下论坛 rolia.net