mersenneforum.org  

Go Back   mersenneforum.org > Prime Search Projects > Conjectures 'R Us

Reply
 
Thread Tools
Old 2008-12-30, 18:57   #23
mdettweiler
A Sunny Moo
 
mdettweiler's Avatar
 
Aug 2007
USA (GMT-5)

792 Posts
Default

Rogue,

I just encountered a very strange error when testing a PRPnet server on localhost loaded with known-prime base 2 Riesel numbers (k*2^n-1). The client grabbed the work, and did the test with LLR as it should (screen output below):
Code:
[2008-12-30 18:48:22 GMT] PRPNET Client application v1.0.0 started
[2008-12-30 18:48:22 GMT] User email address is bugmesticky@googlemail.com
Starting Lucas Lehmer Riesel prime test of 745*2^260463-1
Using Irrational Base DWT : Mersenne fftlen = 14336, Used fftlen = 16384
V1 = 4 ; Computing U0...done.
745*2^260463-1, iteration : 10000 / 260463 [3.83%].
snipped all the iteration counts in here
745*2^260463-1 is prime!  Time : 77.617 sec.
[2008-12-30 18:49:40 GMT] a: Could not find RES64 residue [lresults.txt].  Is llr broken?
And, indeed, there was no lresults.txt file to be found in the PRPnet client directory. Also, the "work_a.out" file looked very strange:
Code:
ABC $a*$b^$c-1
745 2
...it seems to have left the n off completely.

However, when I run the exact same copy of LLR manually on the work_a.in file that PRPnet has left in the client directory, everything looks normal. lresults.txt is present:
Code:
745*2^260463-1 is prime!  Time : 80.790 sec.
...however, work_a.out has still left off the n-value:
Code:
ABC $a*$b^$c-1
745 2
Yet, this apparent error on LLR's part should not affect the lresults.txt file; that still gets produced exactly as it should. Possibly PRPnet is deleting the lresults.txt file before it gets the chance to read it?

Max
mdettweiler is offline   Reply With Quote
Old 2008-12-30, 18:58   #24
nuggetprime
 
nuggetprime's Avatar
 
Mar 2007
Austria

30210 Posts
Default

Quote:
Originally Posted by mdettweiler
Nugget, I presume that the results from email address "hugoplatzer at aon dot at" represent you?
Yes,hugoplatzer AT aon DoT at represents me.
As for the results: They're pretty hard to convert. Maybe I'll learn Perl and write the script, but probably somebody else will faster than me.

BTW,there have no primes been found so far?
And another thing: Can you please increase maxCacheSize to ~1500 per client(not per mailaddress)? My quad isn't connected to the net overnight and can crunch about 1000 WUs per core in twelve hours. Thanks!

nugget

Last fiddled with by nuggetprime on 2008-12-30 at 19:01
nuggetprime is offline   Reply With Quote
Old 2008-12-30, 19:02   #25
nuggetprime
 
nuggetprime's Avatar
 
Mar 2007
Austria

2×151 Posts
Default

Quote:
Originally Posted by mdettweiler View Post
Rogue,
I just encountered a very strange error when testing a PRPnet server on localhost loaded with known-prime base 2 Riesel numbers (k*2^n-1).
<snip>
Max
I'll soon do the same test on my P4.
nuggetprime is offline   Reply With Quote
Old 2008-12-30, 19:02   #26
mdettweiler
A Sunny Moo
 
mdettweiler's Avatar
 
Aug 2007
USA (GMT-5)

792 Posts
Default

Quote:
Originally Posted by nuggetprime View Post
Quote:
Originally Posted by mdettweiler
Nugget, I presume that the results from email address "hugoplatzer at aon dot at" represent you?
Yes,hugoplatzer AT aon DoT at represents me.
As for the results: They're pretty hard to convert. Maybe I'll learn Perl and write the script, but probably somebody else will faster than me.

BTW,there have no primes been found so far?
And another thing: Can you please increase maxCacheSize to ~1500 per client(not per mailaddress)? My quad isn't connected to the net overnight and can crunch about 1000 WUs per core in twelve hours. Thanks!

nugget
First of all: yes, as far as I can tell no PRPs have been found yet. I'll probably try buzzing a known PRP through a local PRPnet server to see what the output file line looks like.

As for the maxCacheSize value: okay, will do.

Last fiddled with by mdettweiler on 2008-12-30 at 19:03
mdettweiler is offline   Reply With Quote
Old 2008-12-30, 19:02   #27
em99010pepe
 
em99010pepe's Avatar
 
Sep 2004

2·5·283 Posts
Default

The IP from the clients should be hidden.
em99010pepe is offline   Reply With Quote
Old 2008-12-30, 19:04   #28
mdettweiler
A Sunny Moo
 
mdettweiler's Avatar
 
Aug 2007
USA (GMT-5)

11000011000012 Posts
Default

Quote:
Originally Posted by em99010pepe View Post
The IP from the clients should be hidden.
You mean in the results file I just posted for Nugget? Hmm...I see what you mean. (I guess I didn't think of that since moderators can see the IP addresses on everyone's forum posts anyway. ) Nugget, if you're done with the results file I posted a few posts back, I'll go ahead and take it offline.
mdettweiler is offline   Reply With Quote
Old 2008-12-30, 19:12   #29
nuggetprime
 
nuggetprime's Avatar
 
Mar 2007
Austria

1001011102 Posts
Default

I downloaded it,you can take it offline.
Can you also tell me how a candidate file for the server should look like?

Last fiddled with by nuggetprime on 2008-12-30 at 19:14
nuggetprime is offline   Reply With Quote
Old 2008-12-30, 19:23   #30
mdettweiler
A Sunny Moo
 
mdettweiler's Avatar
 
Aug 2007
USA (GMT-5)

792 Posts
Default

Quote:
Originally Posted by nuggetprime View Post
I downloaded it,you can take it offline.
Done.
Quote:
Can you also tell me how a candidate file for the server should look like?
Well, the easiest way to make a candidate file is to take an ABC format sieve file, place it in the same directory as an *inactive* server (I'll explain why this is necessary later), and run the command "./prpserver -i abc-file" (where "abc-file" is the name of your sieve file). It will output a candidates file as prpserver.candidates.

Now for why you don't want to do this in the directory of an already-in-use server. That's because when it outputs its new candidates file, rather than appending to any existing ones, it *overwrites* them. So, what I usually do is as follows when I want to add work to a server:

-Generate the candidates file to be added in a separate, "spare" server directory
-Stop the real server
-Concatenate the new candidates file onto the end of the real server's prpserver.candidates file
-Start the real server

This method seems to work pretty well, from what I can tell.
mdettweiler is offline   Reply With Quote
Old 2008-12-30, 19:30   #31
nuggetprime
 
nuggetprime's Avatar
 
Mar 2007
Austria

2·151 Posts
Default

Got a problem: If i setup a local server,i have a problem: The client doesn't get any candidate but reports:INFO:
Code:
No available candidates are left on this server.
It's not completely unable to connect,I've tested that: when i change the port to somewhere where no server is around, it reports another error message:
Code:
Could not connect to any servers and no work is pending
.

nugget
nuggetprime is offline   Reply With Quote
Old 2008-12-30, 19:34   #32
nuggetprime
 
nuggetprime's Avatar
 
Mar 2007
Austria

12E16 Posts
Default

It's now working. I had to put in more candidates.
nuggetprime is offline   Reply With Quote
Old 2008-12-30, 19:37   #33
nuggetprime
 
nuggetprime's Avatar
 
Mar 2007
Austria

12E16 Posts
Default

The "Res64 not found" error also happened to me,also with composite candidates.
nuggetprime is offline   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
PRPNet 5.4.3 Released rogue Software 168 2020-10-10 13:47
PSP goes prpnet ltd Prime Sierpinski Project 86 2012-06-06 02:30
PRPNet 4.0.0 Released rogue Software 84 2011-11-16 21:20
PRPNet 4.0.1 Released Joe O Sierpinski/Riesel Base 5 1 2010-10-22 20:11
PRPNet 3.0.0 Released rogue Conjectures 'R Us 220 2010-10-12 20:48

All times are UTC. The time now is 05:54.

Wed Oct 28 05:54:04 UTC 2020 up 48 days, 3:05, 2 users, load averages: 1.99, 1.65, 1.65

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2020, Jelsoft Enterprises Ltd.

This forum has received and complied with 0 (zero) government requests for information.

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.
A copy of the license is included in the FAQ.