by what you are providing. It looks like you are being overwhelmed by broadcast traffic.
It's called a broadcast storm/multicast storm
http://support.3com.com/infodeli/tools/netmgt/tncsunix/product/091500/c9bstorm.htm
http://support.novell.com/techcenter/articles/nc2000_10c.html
To fix it. I need more details on your infrastructure that runs the network.
Do you use managed switches?
If you do, do you have span-tree protocol running? Do you have a loop?
http://en.wikipedia.org/wiki/Spanning_tree_protocol
I need more details in order to diagnosis your problem further.
But i would use those above links, and do a little searching about broadcast/multicast traffic and what kind of performance problems are caused by this.
Update...
The new updated information is pretty helpful.
Most procurve switches have support for STP (spanning-tree protocol), But you need to configure it properly (assign root bridge, etc). HP calls it RSTP (rapid STP).
this thread may help
http://www.tek-tips.com/viewthread.cfm?qid=1221775&page=1
In order to find the switch. You can use a program called angry ip to scan your ip range
http://www.angryip.org/w/Home
What range you should use depends on what your subnet mask is.
so if you use a class C mask 255.255.255.0 (host should be x.x.x.1-x.x.x.254)
if you use a class b mask 255.255.0.0 (host should be in the range of x.x.1.0-x.x.255.255) it could be huge.
Once you figure out the IP address. You can either access them via a web-broswer or telnet. Telnet will allow more powerful access aka the command line interface (cli)
Post updates if you want to continue this.