|
From: Reed, M. (MREED) <MR...@ar...> - 2004-02-12 21:20:05
|
Someone recently posted a question to this list regarding the wrapper
hanging and not shutting down the wrapped application.
The response was to read integration documentation more closely and to offer
suggestions as to how the docs could be enhanced to help the user solve his
problem. . I have spent a good deal of time reading the docs while
implementing a project using the wrapper.
The wrapper is great stuff, but I have a jew unanswered question that I have
been trying to solve on my own without posting questions that are obviously
answered in the docs.
I have one suggestion or request regarding how the docs could be improved to
help new users or users in general. The integration Method 3 states "The
third and final method, while providing the most flexibility and access to
all of the Wrapper's features"
It would help if a clear distinction of what features are covered by each
method and finally a clear list of the features that one gets by using each
method that you dont get from the others. Specifically what is the list of
features you get using Method 3 that you dont get with Method 1 or Method 2.
Otherwise you study all three, study the example source provided and it a
bit difficult to distinguish what you get from 3 that you dont get from 1 or
2.
I am trying to decide which method is best for my solution which is a fairly
complex server application with lots of redundancy and availability
requirements, remote support requirements via a Web based administration
interface etc......
I have used Methods 1 and Method 3 in some proof of concept code and was not
quite successful with Method 3 due to my lack of understanding I'm sure. A
better understanding of what you get from 3 as oppossed to 1 and 2 would
help hence my intro paragraph.
Question 1:
When using Method 1 and providing my own shut down hook, should I expect to
be able to instantiate an instance of the WrapperActionServer and have it
control my app for example sending an R to restart seems to have no effect
so perhaps this is only usable with Method 3.
Question 2:
When using Method 1 should I expect to be able to configure
wrapper.filter.trigger.1=(some fully qualified exception name)
wrapper.filter.action.1=RESTART
throw this exception from some point in the application, no handle and
therefore cause the Wrapper (using METHOD 1) to restart the service?
Question 3:
What ways can i trigger this mechanism or the wrapper to see my string
filter it and react. For example could I just perform a
System.out("MYCOMMAND") and assuming i had
wrapper.filter.trigger.1=MYCOMMAND
wrapper.filter.action.1=RESTART
configured this would activate restart?
Regards MRR
|