Server data questions

quikdave · 2003-03-22, 00:57

Two questions for one here ...

a) The 'top producers' numbers on the server appear stuck today - the hourly updates are all identical (no new work being credited). I checked down as far as the top 1200, and nobody appears to have done anything (unlikely).

b) Does anyone have any =historic= copies of the full 'top producer' report (3.4Mbytes each!) that they may have downloaded and kept? I'm trying to research the development/growth/history of the primenet server, and some historic data (just maybe one per year) could help a lot.

Alternatively, does anyone know if it is possible to get the server to spit out some more information for each of the users - i.e. when they last submitted a complete unit, or 'ceased trading' or similar? (Right now it is tough to identify a moribund user - someone who switched to SETI, or some other Distributed project).

TIA

trif · 2003-03-22, 01:18

The server is not accepting results right now, which is why no new work has been credited.

It will be fixed soon, I am sure.

NickGlover · 2003-03-22, 03:52

Quote:

Originally Posted by quikdave

Two questions for one here ...
b) Does anyone have any =historic= copies of the full 'top producer' report (3.4Mbytes each!) that they may have downloaded and kept? I'm trying to research the development/growth/history of the primenet server, and some historic data (just maybe one per year) could help a lot.

Here's something from 1999: http://www.juergen-mueller.de/mersenne/

Also, you can use the Internet Archive to get some historic copies of the Primenet top 100 producers (currently http://mersenne.org/ips/tops.shtml and formerly http://entropia.com/ips/tops.shtml). I have not had much luck using the Internet Archive to grab old copies of larger files, such as the full top producers report.

quikdave · 2003-03-22, 12:49

Thanks .. my German is a bit rusty, but I think I can get a couple of useful 1999 data points out of the website. Now if I can just find someone with 2001 data (+/- a bit) I'll be well pleased.

I finally figured out that the server was sick. Strange that it is well enough to keep churning out hourly reports (well it was, even that seems to have ceased now).

guido72 · 2003-03-24, 16:01

Quote:

Originally Posted by quikdave

Two questions for one here ...
b) Does anyone have any =historic= copies of the full 'top producer' report (3.4Mbytes each!) that they may have downloaded and kept? I'm trying to research the development/growth/history of the primenet server, and some historic data (just maybe one per year) could help a lot.
TIA

Hi! I've collected some data from Maj 2001 to Sept 2002. I've put them in a spreadsheet ( roughly 8 Mb, compressed .xls file)... If you think they may interest you, just ask it!
Regards

quikdave · 2003-03-24, 17:34

I've now managed to find (on the internet archive) a more or less complete ('top 26000') list for 12/Apr/01 and 3/Feb/02, however something in the middle of those would be good - how far down the 'top NN' list did you go?

NickGlover · 2003-03-24, 22:43

Quote:

Originally Posted by quikdave

I've now managed to find (on the internet archive) a more or less complete ('top 26000') list for 12/Apr/01 and 3/Feb/02, however something in the middle of those would be good - how far down the 'top NN' list did you go?

Try http://web.archive.org/web/*/http://...roducers.shtml. I was able to get 20/Aug/01, 8/Nov/01, and 18/Dec/01.

Also, http://web.archive.org/web/*/http://...menet/tops.txt will get you one from 1998 if you want to go that far back.

ewmayer · 2003-03-24, 23:42

Quote:

Originally Posted by quikdave

b) Does anyone have any =historic= copies of the full 'top producer' report (3.4Mbytes each!) that they may have downloaded and kept? I'm trying to research the development/growth/history of the primenet server, and some historic data (just maybe one per year) could help a lot.

Also remember that there may be big differences between the primenet TP report and the master TP report ( http://www.mersenne.org/top.htm ), esp. for those users who do lots of manual testing.

quikdave · 2003-03-25, 13:10

Thanks .. yes, I'm aware that the Primenet automatic data is incomplete, however it has the advantage of containing 'average CPU rate' data which allows calculation of when the UID started testing (albeit there are a couple of cases in the database which calculate out to 1969 or 1970!). It is also pretty representative of the 'general mass' of drop-in users .. which tend to be the 'floating voters' who may have moved off to SETI, Cancer Research molecular folding, or whatever.

I'm just looking at the 'top LL testers' data (ignoring, pretty much, factoring). Still looking for some 'full report' data for somewhere in 3q/4q 2002, 1q/2q 1999, and 4q 00.

Obviously data points 'too close together' give misleading work rate info, (because of quatization effect), but they are useful in spotting when a UID dropped out. Would be even more useful if this data could be gotten from the server (or right now, any data!).

2003-03-22, 00:57	#1
quikdave Mar 2003 5 Posts	Server data questions Two questions for one here ... a) The 'top producers' numbers on the server appear stuck today - the hourly updates are all identical (no new work being credited). I checked down as far as the top 1200, and nobody appears to have done anything (unlikely). b) Does anyone have any =historic= copies of the full 'top producer' report (3.4Mbytes each!) that they may have downloaded and kept? I'm trying to research the development/growth/history of the primenet server, and some historic data (just maybe one per year) could help a lot. Alternatively, does anyone know if it is possible to get the server to spit out some more information for each of the users - i.e. when they last submitted a complete unit, or 'ceased trading' or similar? (Right now it is tough to identify a moribund user - someone who switched to SETI, or some other Distributed project). TIA

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
newPGen "Data Execution Prevention" on Windows Server R2 2012	MisterBitcoin	Software	4	2017-02-21 15:50
GPU TF vs DC/LL data	bcp19	GPU to 72	0	2011-12-02 16:41
Data available?	Prime95	LMH > 100M	10	2007-06-22 23:55
New server data functionality	Complex33	Data	4	2003-10-04 00:05
New Server Hardware and price quotes, Funding the server	Angular	PrimeNet	32	2002-12-09 01:12

2003-03-22, 01:18	#2
trif Aug 2002 202₁₀ Posts	The server is not accepting results right now, which is why no new work has been credited. It will be fixed soon, I am sure.

2003-03-22, 12:49	#4
quikdave Mar 2003 5 Posts	Thanks .. my German is a bit rusty, but I think I can get a couple of useful 1999 data points out of the website. Now if I can just find someone with 2001 data (+/- a bit) I'll be well pleased. I finally figured out that the server was sick. Strange that it is well enough to keep churning out hourly reports (well it was, even that seems to have ceased now).

2003-03-24, 17:34	#6
quikdave Mar 2003 5₈ Posts	I've now managed to find (on the internet archive) a more or less complete ('top 26000') list for 12/Apr/01 and 3/Feb/02, however something in the middle of those would be good - how far down the 'top NN' list did you go?

2003-03-25, 13:10	#9
quikdave Mar 2003 5 Posts	Thanks .. yes, I'm aware that the Primenet automatic data is incomplete, however it has the advantage of containing 'average CPU rate' data which allows calculation of when the UID started testing (albeit there are a couple of cases in the database which calculate out to 1969 or 1970!). It is also pretty representative of the 'general mass' of drop-in users .. which tend to be the 'floating voters' who may have moved off to SETI, Cancer Research molecular folding, or whatever. I'm just looking at the 'top LL testers' data (ignoring, pretty much, factoring). Still looking for some 'full report' data for somewhere in 3q/4q 2002, 1q/2q 1999, and 4q 00. Obviously data points 'too close together' give misleading work rate info, (because of quatization effect), but they are useful in spotting when a UID dropped out. Would be even more useful if this data could be gotten from the server (or right now, any data!).