> p.s. Does this discussion really belong on the PrimeNet list rather
> than the mersenne list?
I agree. I'm guessing few Mersenne list participants have a decided interest in
PrimeNet operations issues. Perhaps it suffices to post a gentle reminder of
the PrimeNet list's address, as needed <[EMAIL PROTECTED]>.
> Well, when I had thousands of new machines hitting Primenet back in May, it
> did cause problems...fortunately Scott got on the ball and got it going, and
> it was also related to the 5.2M+ problem, but there were also times (Scott,
> correct me if wrong) that Primenet was already maxed on connections and had
> to refuse other incoming connections.
Old news: Yup, PrimeNet saturated every 15 minutes or so. It has a 64
concurrent clients limit, which normally never gets above 5 or 6, and nominally
hovered around 0.2 back then. I recall something crazy like 20 clients/sec.
More old news: Aaron deployed an insufficiently tested first v16 NTPrime, which
misidentified itself as a v15 FFT client. We got sloppy in the rush to get v16
done because the < 5.26M exponents burned up *much* faster than anticipated.
(v16 Prime95 was well tested.) Work had gone into the server's v16 assignments
code, assuming the v15 client base was a stable stream. I ended up hacking the
server to catch and fix the misidentified packets.
(This is a good example of something for the primenet list.)
Regards,
scott
