Server Overdue/Expire

Complex33 · 2004-03-08, 14:36

I think I might have brought this up before but here are some thoughts. Would it be beneficial if the default time between check-ins for Prime95 was set to a number much lower than 28, say 1 or 2? This would cause the effective overdue/expire time to be reduced from 88 days to around 60 days, not to mention provide better exponent health/status map. I've noticed that the majority of expiring exponents are from accounts that are setup, reserve work, and then never communicate again. I would hope that a change here would reduce the number of stale exponents waiting for reassignment. Also I'd like to have the overdue check in days reduced from 60 but I don't know how many would be in favor of that. I'm just wanting to reduce the band of assigned exponents and try to clean up the trailing edge from exponents that repeatedly get assigned and never worked on. Any thoughts on this?

garo · 2004-03-08, 17:50

This is a proposal worth considering but one also has to keep in mind the fact that some casual participants may be upset if Prime95 tries to communicate with the server too often. one of the great selling points of this project is it's fabulous capacity to work offline.

If the update number is set to one or two days, your casual participant is going to see a message like this very very often:

"ERROR 2250: Cannot contact Primenet Server"

Basically, everytime the machine is not connected to the Internet and Prime95 wants to send an update.

That having been said, we should do something to reduce the impact of these lay accounts on the huge rage in which exponents are assigned these days. The casual account count - measured by the largest Sxxxxxx account assigned - has gone up rapidly recently and is now up to S175000 or so. What's more, my preliminary datamining shows that 90% of these Sxxxxxx accounts never return a result.

PrimeCruncher · 2004-03-08, 22:48

I agree with garo, 1 or 2 days between updates is too often. I think 7 days would be resonable.

jinydu · 2004-03-09, 13:52

Is it possible to view the history of ANY LL test (that is, a graph of iteration number against time)?

MrHappy · 2004-03-09, 14:11

Quote:

Originally Posted by jinydu

Is it possible to view the history of ANY LL test (that is, a graph of iteration number against time)?

GP2 has shown something like that but it's not possible for non-prime95 clients, i think.

2004-03-08, 14:36	#1
Complex33 Aug 2002 Texas 5×31 Posts	Server Overdue/Expire I think I might have brought this up before but here are some thoughts. Would it be beneficial if the default time between check-ins for Prime95 was set to a number much lower than 28, say 1 or 2? This would cause the effective overdue/expire time to be reduced from 88 days to around 60 days, not to mention provide better exponent health/status map. I've noticed that the majority of expiring exponents are from accounts that are setup, reserve work, and then never communicate again. I would hope that a change here would reduce the number of stale exponents waiting for reassignment. Also I'd like to have the overdue check in days reduced from 60 but I don't know how many would be in favor of that. I'm just wanting to reduce the band of assigned exponents and try to clean up the trailing edge from exponents that repeatedly get assigned and never worked on. Any thoughts on this?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
ok to let it expire?	Richard Cameron	Information & Answers	4	2011-09-30 20:17
Isn't a new M. prime overdue?	joblack	Lounge	86	2010-12-28 18:35
Do LL's expire?	hj47	PrimeNet	2	2009-06-14 04:56
Overdue Prime Rally	masser	Sierpinski/Riesel Base 5	49	2008-08-24 20:19
Will my exponent expire?	edorajh	Data	2	2003-11-12 13:08

2004-03-08, 17:50	#2
garo Aug 2002 Termonfeckin, IE 2²·691 Posts	This is a proposal worth considering but one also has to keep in mind the fact that some casual participants may be upset if Prime95 tries to communicate with the server too often. one of the great selling points of this project is it's fabulous capacity to work offline. If the update number is set to one or two days, your casual participant is going to see a message like this very very often: "ERROR 2250: Cannot contact Primenet Server" Basically, everytime the machine is not connected to the Internet and Prime95 wants to send an update. That having been said, we should do something to reduce the impact of these lay accounts on the huge rage in which exponents are assigned these days. The casual account count - measured by the largest Sxxxxxx account assigned - has gone up rapidly recently and is now up to S175000 or so. What's more, my preliminary datamining shows that 90% of these Sxxxxxx accounts never return a result.

2004-03-08, 22:48	#3
PrimeCruncher Sep 2003 Borg HQ, Delta Quadrant 2·3³·13 Posts	I agree with garo, 1 or 2 days between updates is too often. I think 7 days would be resonable.

2004-03-09, 13:52	#4
jinydu Dec 2003 Hopefully Near M48 1758₁₀ Posts	Is it possible to view the history of ANY LL test (that is, a graph of iteration number against time)?