mersenneforum.org PRPnet Servers for CRUS
 Register FAQ Search Today's Posts Mark Forums Read

 2014-09-09, 22:42 #12 MyDogBuster     May 2008 Wilmington, DE 22×23×31 Posts Max, Do we have a CRUS equivalent for this link. http://www.noprimeleftbehind.net/prpnet
2014-09-10, 01:06   #13
mdettweiler
A Sunny Moo

Aug 2007
USA (GMT-5)

186916 Posts

Quote:
 Originally Posted by MyDogBuster Max, Do we have a CRUS equivalent for this link. http://www.noprimeleftbehind.net/prpnet
No, we don't at present. I hadn't made one since the PRPnet server's own stats provide much of the same information. Would you like one?

2014-09-10, 05:39   #14
MyDogBuster

May 2008
Wilmington, DE

22·23·31 Posts

Quote:
 No, we don't at present. I hadn't made one since the PRPnet server's own stats provide much of the same information. Would you like one?
Only if it's not much trouble. That way I can see the whole PRPnet picture without going to multiple screens.

2014-09-10, 07:24   #15
gd_barnes

May 2007
Kansas; USA

25·331 Posts

Quote:
 Originally Posted by rogue Should the sieving posts in this thread be moved to a new thread?
Done. The S9 and general sieving discussion has been moved to a separate thread here.

 2014-09-11, 22:53 #16 mdettweiler A Sunny Moo     Aug 2007 USA (GMT-5) 3×2,083 Posts Gary and/or Ian: could you please update the first post of this thread to reflect the following? This will resolve the "these links will be tweaked later" issue. Thanks. Just delete everything that's in there right now, and paste the contents of the following code block. Code: The following is a list of all of CRUS's active PRPnet servers. The admins will keep this thread updated as often as possible. :smile: [B]Currently processing ranges last updated 2014-09-10 07:30 GMT.[/B] [B]CRUS PRPnet server #1:[/B] Short identification: G1300 server = [b]www.noprimeleftbehind.net[/b] port = 1300 Sierp base 9: 1k remaining to n=2M n-range: 1.5M-2.0M currently processing at n= [COLOR=blue]~1.5M[/COLOR] server summary: [URL]http://noprimeleftbehind.net:1300/all.html[/URL] [B]CRUS PRPnet server #2:[/B] Short identification: G1400 server: [b][url]www.noprimeleftbehind.net[/url][/b] port: 1400 Riesel-Sierp base 2/4 even-odd k-n n-range: 2.0M-5.0M currently processing at n= [COLOR=#0000ff](not started)[/COLOR] server summary: [URL]http://noprimeleftbehind.net:1400/all.html[/URL] Please see the PRPnet server summary pages listed above for current user and team stats. For archived stats from previous CRUS PRPnet drives, see below: [url=http://www.noprimeleftbehind.net/downloads/port%201300%20stats%20as%20of%202014-09-07.pdf]Port 1300 stats as of 2014-09-07[/url] [url=http://www.noprimeleftbehind.net/downloads/port%201400%20stats%20as%20of%202014-09-07.pdf]Port 1400 stats as of 2014-09-07[/url] Since the "old" stats from base 6 (and the earlier 1/2/3/4-k drive) in port 1400 have been archived in the snapshot linked above, I will be clearing them from the server database shortly. This will leave a "clean slate" for the new effort, as we did for port 1300. Last fiddled with by mdettweiler on 2014-09-11 at 22:57 Reason: fix code block
2014-09-11, 23:12   #17
gd_barnes

May 2007
Kansas; USA

25·331 Posts

Quote:
 Originally Posted by mdettweiler Gary and/or Ian: could you please update the first post of this thread to reflect the following? This will resolve the "these links will be tweaked later" issue. Thanks.
Done. Thanks Max! Note that I already cleaned out port 1400 and tweaked the page heading. I saved off the R6 results (residues) to my normal place on my laptop.

Last fiddled with by gd_barnes on 2014-09-11 at 23:13

2014-09-12, 00:07   #18
mdettweiler
A Sunny Moo

Aug 2007
USA (GMT-5)

3·2,083 Posts

Quote:
 Originally Posted by gd_barnes Done. Thanks Max! Note that I already cleaned out port 1400 and tweaked the page heading. I saved off the R6 results (residues) to my normal place on my laptop.
Right. Just so we're clear, I'm talking about clearing out the user/team stats - we had a lot of "crud" pile up there between misspelled team names, computer names accidentally entered as usernames, etc. (If this were at NPLB, we would just enter the incorrect name as an alias, and you'd never know the difference...but without the stats database it's not really convenient to fix.) Hence, my thinking that it'd be good to make a fresh start for the new effort. As noted previously, the old stats have all been archived to the link in the first post.

 2014-09-12, 00:40 #19 mdettweiler A Sunny Moo     Aug 2007 USA (GMT-5) 11000011010012 Posts Port 1400 upgraded to PRPnet 5.3.2 Port 1400 has now been upgraded to the latest PRPnet v5.3.2. It is ready for loading with the base 2/4 candidates as soon as is desired.
 2014-09-12, 03:49 #20 gd_barnes     May 2007 Kansas; USA 25×331 Posts Mark, It looks like there is a teeny nit bug on the PRPnet 5.3.2 port 1400 status page. For the Riesel base 2 k's on the k's remaining line, it says: S2 k's: 10 I think it should say: R2 k's: 10. Perhaps I'm misunderstanding what the "S" means since this is the first look that I have gotten at 5.3.2. Assuming that it is an issue, it might have happened because I loaded it in 2 separate steps. I had two files and so loaded the 10 Riesel k's first followed by loading the 5 Sierp k's. Gary
 2014-09-12, 12:56 #21 mdettweiler A Sunny Moo     Aug 2007 USA (GMT-5) 3·2,083 Posts Ah yes, I just noticed that myself yesterday. The relevant code seems to be here, lines 83-89 of SierpinskiRieselHTML.cpp (and again in an identical section at the end of the loop, lines 183-189): Code:  if (ib_ServerStatsSummaryOnly) ip_Socket->Send("
%c%d k\'s: %d", (c > 0 ? 'S' : 'R'), prevB, conjectureKs); else ip_Socket->Send("
%c%d k\'s: %d", (conjectureKs == conjecturePRPsAndPrimesFound ? "red" : "yellow"), (c > 0 ? 'S' : 'R'), prevB, conjectureKs); In this case, we're taking the "else" branch since we have the server configured to display detailed info for all bases, not just a single summary. Actually, the code seems to be correct - in fact it's exactly what I was planning to fix it to when I opened the code to find the bug. This leaves me rather puzzled...but, that's probably just my acolyte-grade knowledge of the PRPnet codebase showing. Mark, do you know if there's anywhere else in the code that could be generating or overriding this section of HTML for a Sierpinski/Riesel conjecture server?
 2014-09-12, 13:09 #22 mdettweiler A Sunny Moo     Aug 2007 USA (GMT-5) 186916 Posts On a second read-through, I found the bug: it was staring me in the face the whole time. You'll notice that the "Send()" lines are making reference to "prevB" - that's because they're trigger upon reading the first row from the database that doesn't match the base of the table we're building, thus signifying that we're done with that base and should cap it off with a summary row. So, "prevB" and "prevC" contain the actual base and c value that applies to this line on the web page. This was causing each table summary to refer to the next base's Sierpinski/Rieselness. The last base displayed correctly, because there were no more rows to read from the DB, which pushed it to the code at lines 183-189 outside the loop, where "c" and "prevC" are the same value. So, the code should really look like this: Code:  if (ib_ServerStatsSummaryOnly) ip_Socket->Send("
%c%d k\'s: %d", (prevC > 0 ? 'S' : 'R'), prevB, conjectureKs); else ip_Socket->Send("
%c%d k\'s: %d", (conjectureKs == conjecturePRPsAndPrimesFound ? "red" : "yellow"), (prevC > 0 ? 'S' : 'R'), prevB, conjectureKs); I'll send a diff to Mark so he can include it in the next release. In the meantime, it's fixed on our servers. Last fiddled with by mdettweiler on 2014-09-12 at 13:19

 Similar Threads Thread Thread Starter Forum Replies Last Post mdettweiler No Prime Left Behind 228 2018-12-26 04:50 rogue Open Projects 26 2013-01-16 01:33 Joe O Sierpinski/Riesel Base 5 6 2010-12-06 20:41 opyrt Prime Sierpinski Project 13 2009-11-04 21:33 gd_barnes Conjectures 'R Us 39 2008-07-15 10:26

All times are UTC. The time now is 19:42.

Fri Jan 21 19:42:23 UTC 2022 up 182 days, 14:11, 0 users, load averages: 1.82, 1.44, 1.29