Wait on I just realized it shouldn't be cutting out at all.
My original problem was I thought when it was totally locking up after 10-20 mins, was that it was staying that way, but now I see 30 secs later it gets going again. When it stops and its actually every 19 minutes it says:
INFO proxy getwork_listner._on_authorised # "My Miners Name" 'asks for new work.
Im using a wireless 4 port modem router,but im not on wifi, my pc is on one of the Ethernet ports and the Blade is on another, leaving 2 spare of course. I might try turning off the wireless signal and see what happens.
That is almost the *exact* configuration I started out with when I experienced problems. It took about a week for me to figure out that I needed to isolate the boards on their own network. At the high point, I had 10 blades on a 16 port switch. Now I have 3 (sold the rest), but like I said... I no longer have any problems.
As best as I can illustrate, my configuration looks like:
INTERNET<--->modem<--->(eth0)[computer-running-bfgminer]
(eth1)<--->netgear-ethernet-switch
^ ^
| |
v v
blade0 ... bladeN
I originally went down this path to try to isolate my network issues in the event that something
was either broadcasting or interfering with the other hosts on its subnet. Basically, all of the
blades are on a 192.168.2.x subnet along with eth1. The computer running bfgminer is on my
internal internet connected network 192.168.1.x. This means that bfgminer running in proxy mode can get out to communicate with whichever pool you've assigned it to. The blades themselves don't need to access the internet directly, they only need to talk to the host running bfgminer.
Since switching to this configuration, the remaining boards have been running without incident. They've even been stable enough that I have developed tools to monitor, configure, and reconfigure them automatically. (I'll probably be releasing the tools for free to the public under a GPL license within the next couple of weeks).
MP