O
oddaolsen
Hi.
I need to build some utilities which are going to be used to monitor
our distributed applications. We have (the quite common) scenario of a
highly distributed environment, with an EJB layer acting as an
integration layer against very many different backends.
The different applications/backends have a very variable degree of
self detection and error handling. The consequence of this is of
course that it is difficult to monitor all the different elements, it
is difficult to do error searching and diagnostics and so on.
The best approach would of course be to improve the applications
themselves. But the applications are delivered from different vendors,
they (the backends) are on different technologies and so on.
What I need is a simple framework that enables me to implement
monitoring services bit by bit. The services need to be implemented in
a way (threads) that gives me the ability to do a time-out if the
request to a specific backend system hangs. The framework must give me
the ability to specify/configure wich monitoring services to run, I
must be able to specify what test data to use, to monitor response
time, to be able to compare response time to threshold values, and so
on.
I do not want to write this framework if a good (open source) one
already exists. Junit may be one option. I've tried to search for
other ones. Found some but I am not yet satisfied.
Does anyone have something to recommend?
I need to build some utilities which are going to be used to monitor
our distributed applications. We have (the quite common) scenario of a
highly distributed environment, with an EJB layer acting as an
integration layer against very many different backends.
The different applications/backends have a very variable degree of
self detection and error handling. The consequence of this is of
course that it is difficult to monitor all the different elements, it
is difficult to do error searching and diagnostics and so on.
The best approach would of course be to improve the applications
themselves. But the applications are delivered from different vendors,
they (the backends) are on different technologies and so on.
What I need is a simple framework that enables me to implement
monitoring services bit by bit. The services need to be implemented in
a way (threads) that gives me the ability to do a time-out if the
request to a specific backend system hangs. The framework must give me
the ability to specify/configure wich monitoring services to run, I
must be able to specify what test data to use, to monitor response
time, to be able to compare response time to threshold values, and so
on.
I do not want to write this framework if a good (open source) one
already exists. Junit may be one option. I've tried to search for
other ones. Found some but I am not yet satisfied.
Does anyone have something to recommend?