Re: Unexpected low CPU usage

> I didn't find anything there.

Sorry, I forgot about the limit of maximum 10 downloads. I now uploaded it
here:
http://www.files-up.com/129WT7NJRPDT/CPU-utilization.png.html

> However - one core is fully loaded - okay, that really points to a big
> point, doesn't it? Is the CLIENT multi-threaded?

No, none of the clients are.

> Why don't you just fire up 100 clients, and see if the box goes to 100%
> then? Why not load it both internally and externally (from another
> physical box), preferably with _several_ external boxes? That's my idea
> of fun, at least: load it till DIES, and try to pull some information
> out of that.

I will have to do those extreme load tests too because I will run into
other problems then. One problem I encountered with the RMI client is the
exhaustion of free ports under Windows. I will need to modify my socket
factories so they accept a port range.

What I don't understand is that a protocol that is lightweight when it
comes to serialization of the parameters and results, can keep one of my
cores busy (indicating it is not disk-/network-bound), but when I switch
to a protocol that apparently requires more CPU power for serialization
the CPU is no longer used fully.

> Again: if the client (or server) is "single threaded", or at least
> single-access in some way (it waits until reply before posting a new
> request), you will of course never get two CPUs loaded 100% - you'll
> maybe even have a hard time filling one CPU, since there will inevitably
> be _some_ IO bound (not-CPU-loading, e.g. lag) elements in there per
> request/response round trip. (Or maybe not if this is done fully in the
> OS, that is, nothing goes out on any wire, and hence there aren't
> actually any CPU-waiting moments in a request/response roundtrip - but
> that will just fill one CPU at any rate).

In my earlier posts when I mentioned 'CPU' I meant a single core CPU. When
I posted the CPU graphs, they were of course not showing 100 % usage
because I have two cores. I realize I can never get my dual core CPU fully
utilized with a single-threaded client.

I get one of my cores fully utilized using RMI and XML-RPC and I don't
understand why this is not happening for SOAP.

> You didn't answer the question (or do what I told you there)!

I am sorry, but I think my issue lies elsewhere. I have a CPU intensive
protocol which is not fully utilizing the CPU. I am just trying to
understand what the underlying issue can be? I have looked at locking and
synchronization, background file access, background network access, etc.

> And btw, here you even state "single threaded" - is it really this
> obvious?

Do you mean that I would never be able to get one core fully utilized with
a single thread? If so, see above.

Johan

===================================
This list is hosted by DevelopMentor®  http://www.develop.com

View archives and manage your subscription(s) at http://discuss.develop.com