My computer, Dual p3, is set with a connection time of one(1) day. My resource share is Einstein 100, Seti 250 normally.
Due to the latest communication problem I almost ran out of work for Seti, but no problem it downloaded two more units from Einstein. I thought OK when Seti is back on

it will recognise I have sufficient work and it will only download, at most, a couple of Seti units. But no it down loaded the usual amount.

I now have 4 Einstein, 2 processing
and 6 Seti again 2 processing

Boinc manager, due to the inaccurate Benchmarks gives the estimate of processing as Einstein - 14:27 and Seti as 10:05. This would make (4 * 14.5) + (6 * 10) = 118hrs

for two processors i.e 59 hours of work minus of course the work done, which is a total of 28:40 approx. Therefore each processor has over 44 hours of work to

complete, on the Managers estimate.

The actual times for processing are Einstein 19:30 and Seti 8:20 on a quick average calculation so that the calculation per processor is ((4 * 19.5) + (6 * 8.33)) / 2 = 64 hrs

minus the processing done 14.33 = 49.67 hours.

As I don't think this is how it was intended to work I have adjusted the resource share to 100:100 so that the Einstein unts have a fair chance of completeing in the 7 day

time frame without going into panic mode, although I think it still will on the 2 * connection time rule.

Comments please from those who really know if this how its meant to work, even though I suspect this is going to be over criticised, hard working JM7 again. And a further

request, put the benchmarking problem to the top of the list immediately, I don't believe that the scheduling is going to work correctly until it is.

Regards

Andy