After sevral months of trouble free operation, I ended up wasting close to 4 hours this afternoon trying to get repo processing done. So it looks as though what I thought was the solution must have conencided with hardware fixes at the remote end. Problem seems pretty much exactly what it was before - attempting a long processing run results in a locked up session. Attempting to reconnect will usually time out, and when I do get a connection, running uptime shows a machine uptime of only a few minutes.
I'll pass the problem on to support, but that didn't get me far last time, so I'm not holding my breath for it this time.
I simply do not have the time or patience to keep trying over and over again. I will report the problem, but until such time as support tells me it is fixed, I will not even attempt to process any updates.
Wonderful timing, with VL7.0 in the wings.....