| Author |
Message |
|---|
|
|
|
Discussion about PRPNet >= 4.3.6
I have done our multicore archive here: http://uwin.mine.nu/PRPNet/ |
|
|
|
|
|
[2011-11-19 12:47:55 MZ] GFN262144: PRPNet server is version 4.3.5
Hi Wellcome to PrimeGrid prpserver SR5 server
GeneferCUDA 0.97 (CUDA3.2) based on Genefer v2.2.0
Copyright (C) 2001-2003, Yves Gallot (v1.3)
Copyright (C) 2009, Mark Rodenkirch, David Underbakke (v2.2.0)
Copyright (C) 2010, Shoichiro Yamada (CUDA)
A program for finding large probable generalized Fermat primes.
Start test of file 'work_GFN262144.in' - 12:47:55
Does that look good ??
I mean is this the latest version running on my graphics card ??
Best regards + tks
Parabol
____________
I'm a prime millionaire !
9*2^3497442+1 |
|
|
John Honorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
                 
|
[2011-11-19 12:47:55 MZ] GFN262144: PRPNet server is version 4.3.5
Hi Wellcome to PrimeGrid prpserver SR5 server
GeneferCUDA 0.97 (CUDA3.2) based on Genefer v2.2.0
Copyright (C) 2001-2003, Yves Gallot (v1.3)
Copyright (C) 2009, Mark Rodenkirch, David Underbakke (v2.2.0)
Copyright (C) 2010, Shoichiro Yamada (CUDA)
A program for finding large probable generalized Fermat primes.
Start test of file 'work_GFN262144.in' - 12:47:55
Does that look good ??
I mean is this the latest version running on my graphics card ??
Yes! :)
____________
|
|
|
|
|
[2011-11-19 12:47:55 MZ] GFN262144: PRPNet server is version 4.3.5
Hi Wellcome to PrimeGrid prpserver SR5 server
GeneferCUDA 0.97 (CUDA3.2) based on Genefer v2.2.0
Copyright (C) 2001-2003, Yves Gallot (v1.3)
Copyright (C) 2009, Mark Rodenkirch, David Underbakke (v2.2.0)
Copyright (C) 2010, Shoichiro Yamada (CUDA)
A program for finding large probable generalized Fermat primes.
Start test of file 'work_GFN262144.in' - 12:47:55
Does that look good ??
I mean is this the latest version running on my graphics card ??
Yes! :)
Thanks John !
It was just by chance that I installed my new Graphics-Card at the time
of Michael's great discovery !! Congrat !!
As you can see I just do some testing on my new rig on PSA.
So far all went well and I installed the files for 4 cores.
However I had issues to connect to different server's.
After some hours of investigation I found this line in the master_client file:
//server=SR5:0:1:prpnet.mine.nu:7171server=PPSElow:100:5:prpnet.mine.nu:12000
After 'uncommenting' between ...7171server=... all went as expected.
I have no Idea why this happened during installation..but maybe an issue
on the downloaded files ??
SomeOne same experience ??
Tks + happy crunching
Parabol
____________
I'm a prime millionaire !
9*2^3497442+1 |
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
I have no Idea why this happened during installation..but maybe an issue
on the downloaded files ??
Yes, issue with download files. I must remember to use a text editor that will save CRLF for Windows.
____________
|
|
|
|
|
|
OK and thanks again !
Maybe I missed something reading the instruction files ?
Or could that information take place somewhere in there ?
I know reading about it somewhere in the Forum...
..but I'm lazy and not so practiced to computer...
Parabol
____________
I'm a prime millionaire !
9*2^3497442+1 |
|
|
|
|
|
Just installed the Linux 64 version of PRPNET 4.3.7, and when I try starting it, the terminal windows won't even open. Anyone else having this issue?
____________
 |
|
|
|
|
Just installed the Linux 64 version of PRPNET 4.3.7, and when I try starting it, the terminal windows won't even open. Anyone else having this issue?
Install xterm.
____________
 |
|
|
|
|
Just installed the Linux 64 version of PRPNET 4.3.7, and when I try starting it, the terminal windows won't even open. Anyone else having this issue?
Install xterm.
Do you see anything in the log ?
Lennart |
|
|
 NeoVolunteer tester
Send message
Joined: 28 Oct 10
Posts: 710
ID: 71509
Credit: 91,178,992
RAC: 0
                  
|
Just installed the Linux 64 version of PRPNET 4.3.7, and when I try starting it, the terminal windows won't even open. Anyone else having this issue?
I think there's a start file in each of your prpclient directories.
Look for it and run it in a terminal.
Neo |
|
|
|
|
|
Does the new prpclient-5.0.2-linux_64-gpu require cuda 4?
I installed 5.0.2 and now I'm getting this error and did not before with the 4.3.6 client:
./GeneferCUDA: error while loading shared libraries: libcudart.so.4: cannot open shared object file: No such file or directory
____________
|
|
|
|
|
Does the new prpclient-5.0.2-linux_64-gpu require cuda 4?
I installed 5.0.2 and now I'm getting this error and did not before with the 4.3.6 client:
./GeneferCUDA: error while loading shared libraries: libcudart.so.4: cannot open shared object file: No such file or directory
I will answer my own question by saying...yes it does. Installed cuda 4 and it took right off! :)
____________
|
|
|
|
|
|
I have a suggestion: Is it possible to add an info line in prpnet client of how much work is left on the port? |
|
|
rogueVolunteer developer
Send message
Joined: 8 Sep 07
Posts: 1259
ID: 12001
Credit: 18,565,548
RAC: 0
|
I have a suggestion: Is it possible to add an info line in prpnet client of how much work is left on the port?
If you mean like server_stats.html shows, then probably not. |
|
|
|
|
|
Hi,
do you think you can add an option (in startoption or stopasoption) to "Complete all work units and Shut Down" ?
I have computers behind proxy that don't have access to prpnet server.
So I manually update the buffer for prpclient.
But I don't want the clients to try to connect to the servers. Just shutdown when they have completed all work units.
It will be great.
Thanks
____________
Badge Score: 1*5 + 8*6 + 8*7 + 3*8 + 1*9 = 142 |
|
|
rogueVolunteer developer
Send message
Joined: 8 Sep 07
Posts: 1259
ID: 12001
Credit: 18,565,548
RAC: 0
|
do you think you can add an option (in startoption or stopasoption) to "Complete all work units and Shut Down" ?
I have computers behind proxy that don't have access to prpnet server.
So I manually update the buffer for prpclient.
But I don't want the clients to try to connect to the servers. Just shutdown when they have completed all work units.
Done. |
|
|
|
|
|
whow... Efficient...
Thanks a lot.
____________
Badge Score: 1*5 + 8*6 + 8*7 + 3*8 + 1*9 = 142 |
|
|
|
|
|
is this possible to normalize servers share percentage automatically?
I mean make them 100% total from the numbers that is entered in config.
>>> You can make any combinations you like. Just make sure ALL NUMBERS ADD UP TO 100. :)
this is thing that require 15 minutes from programmer to fix once and forever.
____________
wbr, Me. Dead J. Dona
 |
|
|
rogueVolunteer developer
Send message
Joined: 8 Sep 07
Posts: 1259
ID: 12001
Credit: 18,565,548
RAC: 0
|
is this possible to normalize servers share percentage automatically?
I mean make them 100% total from the numbers that is entered in config.
>>> You can make any combinations you like. Just make sure ALL NUMBERS ADD UP TO 100. :)
this is thing that require 15 minutes from programmer to fix once and forever.
Done. |
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
is this possible to normalize servers share percentage automatically?
I mean make them 100% total from the numbers that is entered in config.
>>> You can make any combinations you like. Just make sure ALL NUMBERS ADD UP TO 100. :)
this is thing that require 15 minutes from programmer to fix once and forever.
Done.
Wow, nice feature update. Thanks for the suggestion D.
____________
|
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 14036
ID: 53948
Credit: 475,998,733
RAC: 222,579
                             
|
is this possible to normalize servers share percentage automatically?
I mean make them 100% total from the numbers that is entered in config.
>>> You can make any combinations you like. Just make sure ALL NUMBERS ADD UP TO 100. :)
this is thing that require 15 minutes from programmer to fix once and forever.
Done.
Wow, nice feature update. Thanks for the suggestion D.
LOL, want to hear something funny?
I wrote exactly that kind of normalizer code for a system I wrote way back in 1985. In particular, I wrote it so that if someone wanted to split their allocations into three equal parts, they could do 33, 33, and 33 rather than 33, 33, and 34. It is, as Death said, very easy to impliment. This was a financial system, so that little difference between 33 and 34 was important.
That computer system was in production from 1985 until 2008.
I don't think that feature, allowing the percentages to total something other than 100%, was used even once in the 23 years that system was in use.
Go figure. ;-)
____________
My lucky number is 75898524288+1 |
|
|
|
|
|
omg. i can't imagine the size of documentation to update.........
LOL
----
thank you guys ^_^
____________
wbr, Me. Dead J. Dona
|
|
|
|
|
|
I have a big problem with PRPnet 5.0.2. I have terminated windows with CTRL+C. Then I restarted the computer. It looks like there is an incompatibility or a big bug with server 4.3.6. After starting PRPnet again results were returned to server then the app crashed. I tried this a couple of times, nothing. I updated client.exe to the 4.x version and its working. This test was made with the PPSElow2 server.
Win7 64bit, 16GB RAM, GTX460, i5-2500k@4Ghz, temps ok
Update:
I have done a second test with running out all workunits, exit, restart with 5 only, terminated after the first one then restart with success. I have no idea whats wrong. |
|
|
rroonnaallddVolunteer developer
Volunteer tester
Send message
Joined: 3 Jul 09
Posts: 1213
ID: 42893
Credit: 34,634,263
RAC: 0
            
|
|
Do you use an antivirus- or inetsecurity-solution?
____________
Best wishes. Knowledge is power. by jjwhalen
 |
|
|
|
|
Do you use an antivirus- or inetsecurity-solution?
Yes but excluded. |
|
|
rroonnaallddVolunteer developer
Volunteer tester
Send message
Joined: 3 Jul 09
Posts: 1213
ID: 42893
Credit: 34,634,263
RAC: 0
|
Do you use an antivirus- or inetsecurity-solution?
Yes but excluded.
The new executable too?
Some products use a hash-value internally to identify an entry from the exception list. The app-name could be the same but their hash will be different.
____________
Best wishes. Knowledge is power. by jjwhalen
|
|
|
|
|
Do you use an antivirus- or inetsecurity-solution?
Yes but excluded.
The new executable too?
Some products use a hash-value internally to identify an entry from the exception list. The app-name could be the same but their hash will be different.
That was only the 5.0.2 client.exe file. Iam using Norton Internet Security and its not allowed to check/scan/autoprotect this directory. But now its working as it should again. |
|
|
|
|
Added a MacIntel_AVX build provided by Iain Bethune.
If someone could test it and provide feedback, that would be very much appreciated.
I am running an intel iMac, and will give it a go...
I installed the new ver. of prpnet 5.0.3, and enabled my "geneferCUDA32", I get the message:
Last login: Mon Jan 9 16:13:52 on ttys000
ronald-littlefields-imac:~ ronaldlittlefield$ /Users/ronaldlittlefield/Desktop/prpclient-5.0.3-macintel/prpclient-2/start.command ; exit;
Could not find executable './GeneferCUDA32'.
I checked my folders, prpclient-1 & prpclient-2, and there is an .exe with the name "GeneferCUDAMacIntel32.exe" should the master_prpclient.ini file be changed to reflect the new file name? Thoughts...?
____________
http://stats.free-dc.org/pgridtag.php?id=51592&theme=2 |
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
I installed the new ver. of prpnet 5.0.3, and enabled my "geneferCUDA32", I get the message:
Last login: Mon Jan 9 16:13:52 on ttys000
ronald-littlefields-imac:~ ronaldlittlefield$ /Users/ronaldlittlefield/Desktop/prpclient-5.0.3-macintel/prpclient-2/start.command ; exit;
Could not find executable './GeneferCUDA32'.
I checked my folders, prpclient-1 & prpclient-2, and there is an .exe with the name "GeneferCUDAMacIntel32.exe" should the master_prpclient.ini file be changed to reflect the new file name? Thoughts...?
Yes. master_prpclient.ini file for MacIntel should have the following:
//geneferexe=./GeneferCUDAMacIntel32
//geneferexe=./GeneferCUDAMacIntel64 Or, if you wish, you can just change the file names to match what's in the ini file. Either way is fine. Probably quicker to change ini file and hit update. :)
____________
|
|
|
|
|
I installed the new ver. of prpnet 5.0.3, and enabled my "geneferCUDA32", I get the message:
Last login: Mon Jan 9 16:13:52 on ttys000
ronald-littlefields-imac:~ ronaldlittlefield$ /Users/ronaldlittlefield/Desktop/prpclient-5.0.3-macintel/prpclient-2/start.command ; exit;
Could not find executable './GeneferCUDA32'.
I checked my folders, prpclient-1 & prpclient-2, and there is an .exe with the name "GeneferCUDAMacIntel32.exe" should the master_prpclient.ini file be changed to reflect the new file name? Thoughts...?
Yes. master_prpclient.ini file for MacIntel should have the following:
//geneferexe=./GeneferCUDAMacIntel32
//geneferexe=./GeneferCUDAMacIntel64 Or, if you wish, you can just change the file names to match what's in the ini file. Either way is fine. Probably quicker to change ini file and hit update. :)
OK! I'll do that, but not tonight! I gotta hit the hay. Will try tomorrow, and report results... Thank you!
____________
http://stats.free-dc.org/pgridtag.php?id=51592&theme=2 |
|
|
|
|
I installed the new ver. of prpnet 5.0.3, and enabled my "geneferCUDA32", I get the message:
Last login: Mon Jan 9 16:13:52 on ttys000
ronald-littlefields-imac:~ ronaldlittlefield$ /Users/ronaldlittlefield/Desktop/prpclient-5.0.3-macintel/prpclient-2/start.command ; exit;
Could not find executable './GeneferCUDA32'.
I checked my folders, prpclient-1 & prpclient-2, and there is an .exe with the name "GeneferCUDAMacIntel32.exe" should the master_prpclient.ini file be changed to reflect the new file name? Thoughts...?
Yes. master_prpclient.ini file for MacIntel should have the following:
//geneferexe=./GeneferCUDAMacIntel32
//geneferexe=./GeneferCUDAMacIntel64 Or, if you wish, you can just change the file names to match what's in the ini file. Either way is fine. Probably quicker to change ini file and hit update. :)
Ok, I changed the filenames in each folder, but get a different error message that reads:
ronald-littlefields-imac:~ ronaldlittlefield$ /Users/ronaldlittlefield/Desktop/prpclient-5.0.3-macintel/prpclient-2/start.command ; exit;
dyld: Library not loaded: @rpath/libcudart.dylib
Referenced from: /Users/ronaldlittlefield/Desktop/prpclient-5.0.3-macintel/prpclient-2/./GeneferCUDAMacIntel32
Reason: image not found
sh: line 1: 814 Trace/BPT trap ./GeneferCUDAMacIntel32 -v > a.out
Unable to determine version from './GeneferCUDAMacIntel32'
Any ideas?? Thanks in advance!
____________
http://stats.free-dc.org/pgridtag.php?id=51592&theme=2 |
|
|
|
|
|
New start/stop add in 5.0.3 are not available in master_prpclient.ini
Added startoption/stopoption 7, to complete all work then shut down without reporting completed work.
Until the file is modified, will it work if I add manually these options in the ini file?
PS: I only download the linux & Windows 64bit archive
____________
Badge Score: 1*5 + 8*6 + 8*7 + 3*8 + 1*9 = 142 |
|
|
|
|
|
John, will you resubmit all numbers from 503 version?
Now i understand why i found one client hanged at the morning.
____________
wbr, Me. Dead J. Dona
|
|
|
|
|
New start/stop add in 5.0.3 are not available in master_prpclient.ini
Added startoption/stopoption 7, to complete all work then shut down without reporting completed work.
Another question: Does this option also available for stopasapoption?
Or it's already stopasapoption and not stopoption?
If not, could you add it also for stopasapoption?
Thanks
____________
Badge Score: 1*5 + 8*6 + 8*7 + 3*8 + 1*9 = 142 |
|
|
|
|
dyld: Library not loaded: @rpath/libcudart.dylib
You need to install the CUDA SDK from Nvidia. And (if I remember) you might also have to add the /usr/local/cuda/lib directory to your LD_LIBRARY_PATH variable. You can do this automatically by adding the line:
export LD_LIBRARY_PATH=/usr/local/cuda/lib:$LD_LIBRARY_PATH
to your ~/.profile file
Cheers
- Iain[/code]
____________
Twitter: IainBethune
Proud member of team "Aggie The Pew". Go Aggie!
3073428256125*2^1290000-1 is Prime! |
|
|
|
|
|
llravx.exe - what's this???
____________
wbr, Me. Dead J. Dona
|
|
|
HonzaVolunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1963
ID: 352
Credit: 6,398,257,295
RAC: 2,430,869
                                  
|
llravx.exe - what's this???
Check news on homepage or use Search function.
Or check AVX build of llr (20%-50% faster) for example
____________
My stats |
|
|
|
|
|
I've checked and searched.
Gary Craig, a member of Aggie The Pew, was successful in compiling an AVX version of LLR using gwnum v27.2 and LLR v3.8.6dev. We've been testing it for the past week with success. Depending on the LLR project, speed improvements range from 20% to 50%.
my reaction is WOW! COOL!!! WTF IS AVX???
____________
wbr, Me. Dead J. Dona
|
|
|
Scott BrownVolunteer moderator
Project administrator
Volunteer tester
Project scientist
Send message
Joined: 17 Oct 05
Posts: 2416
ID: 1178
Credit: 19,978,285,056
RAC: 19,305,139
                                            
|
my reaction is WOW! COOL!!! WTF IS AVX???
See here.
____________
141941*2^4299438-1 is prime!
 |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 14036
ID: 53948
Credit: 475,998,733
RAC: 222,579
|
my reaction is WOW! COOL!!! WTF IS AVX???
See here.
Less technical answer: AVX is a set of new instructions in the lastest Intel (Sandy Bridge) and AMD (Bulldozer) CPUs which are extremely beneficial to programs such as LLR, if, and only if, those programs are re-written specifically to take advantage of the new CPU instructions.
People have been reporting speed improvements of as much as 50% by switching to the AVX-enabled version of LLR.
____________
My lucky number is 75898524288+1 |
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
New start/stop add in 5.0.3 are not available in master_prpclient.ini
Added startoption/stopoption 7, to complete all work then shut down without reporting completed work.
Until the file is modified, will it work if I add manually these options in the ini file?
PS: I only download the linux & Windows 64bit archive
Yes, they will work. I updated the master_prpclient.ini files in all the archives. Now they are all consistent. Thanks for the catch.
Another question: Does this option also available for stopasapoption?
Or it's already stopasapoption and not stopoption?
Yes, available in all 3 sections.
____________
|
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
John, will you resubmit all numbers from 503 version?
Now i understand why i found one client hanged at the morning.
Please see your post and answer here: http://www.primegrid.com/forum_thread.php?id=3890&nowrap=true#46634
____________
|
|
|
|
|
my reaction is WOW! COOL!!! WTF IS AVX???
See here.
Less technical answer: AVX is a set of new instructions in the lastest Intel (Sandy Bridge) and AMD (Bulldozer) CPUs which are extremely beneficial to programs such as LLR, if, and only if, those programs are re-written specifically to take advantage of the new CPU instructions.
People have been reporting speed improvements of as much as 50% by switching to the AVX-enabled version of LLR.
Yes I can confirm about 50% faster:
[2012-01-11 15:10:39 MZ] Server: PPSElow2, Candidate: 1271*2^298073+1 Program: llravx.exe Residue: 826F1623EF515F13 Time: 32 seconds
Rig: own-build i7-2600K , GTX 560 , Win7 prof SP1
..and a lot of reading in this Forum
This step is awsome ! A great 'thank you' to all who made this possible !!!
Looking forward to help crunching..and of course to find 'that' prime.
Greetings .. parabol
____________
I'm a prime millionaire !
9*2^3497442+1 |
|
|
|
|
|
I guess these are simple questions but others might have it.
How can we tell if our system is a SB (Intel) or Bulldozer (AMD)?
Will the new llravx.exe work normally on an older system (as in this will replace llr.exe someday)? |
|
|
HonzaVolunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1963
ID: 352
Credit: 6,398,257,295
RAC: 2,430,869
|
I guess these are simple questions but others might have it.
How can we tell if our system is a SB (Intel) or Bulldozer (AMD)?
Use CPU-Z for example - http://www.cpuid.com/softwares/cpu-z.html
Will the new llravx.exe work normally on an older system (as in this will replace llr.exe someday)?
Yes.
____________
My stats |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 14036
ID: 53948
Credit: 475,998,733
RAC: 222,579
|
How can we tell if our system is a SB (Intel) or Bulldozer (AMD)?
Don't know about AMD, but I believe the model number of the Intel Core i3, Core i5, and Core i7 chip will tell you which generation chip it is.
First gen chips have a 3 digit number (i.e., Core i7-920), while 2nd gen Sandy Bridge chips have a 4 digit number (i.e., Core i7-2600K).
Also, if you look at the beginning of your BOINC log (CTRL-SHIFT-E on the newer boinc clients), you'll see several lines similar to this:
1/10/2012 9:21:32 PM | | Processor: 4 GenuineIntel Intel(R) Core(TM)2 Quad CPU @ 2.40GHz [Family 6 Model 15 Stepping 7]
1/10/2012 9:21:32 PM | | Processor: 4.00 MB cache
1/10/2012 9:21:32 PM | | Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss htt tm pni ssse3 cx16 syscall nx lm vmx tm2 pbe
1/10/2012 9:21:32 PM | | OS: Microsoft Windows 7: Professional x64 Edition, Service Pack 1, (06.01.7601.00)
1/10/2012 9:21:32 PM | | Memory: 8.00 GB physical, 16.00 GB virtual
1/10/2012 9:21:32 PM | | Disk: 931.51 GB total, 373.21 GB free
1/10/2012 9:21:32 PM | | Local time is UTC -5 hours
1/10/2012 9:21:32 PM | | NVIDIA GPU 0: GeForce GTX 460 (driver version 28586, CUDA version 4010, compute capability 2.1, 1024MB, 605 GFLOPS peak)
That line lists all the features the cpu has. CPUs with AVX instructions should should have AVX (or something like that) on that line.
____________
My lucky number is 75898524288+1 |
|
|
HonzaVolunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1963
ID: 352
Credit: 6,398,257,295
RAC: 2,430,869
|
That line lists all the features the cpu has. CPUs with AVX instructions should should have AVX (or something like that) on that line.
I wonder which one represents AVX.
11/01/2012 13:05:47 | | Processor: 4 GenuineIntel Intel(R) Core(TM) i5-2500 CPU @ 3.30GHz [Family 6 Model 42 Stepping 7]
11/01/2012 13:05:47 | | Processor: 256.00 KB cache
11/01/2012 13:05:47 | | Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss htt tm pni ssse3 cx16 sse4_1 sse4_2 syscall nx lm vmx smx tm2 popcnt aes pbe
11/01/2012 13:05:47 | | OS: Microsoft Windows Server 2008 "R2": Standard x64 Edition, Service Pack 1, (06.01.7601.00)
11/01/2012 13:05:47 | | Memory: 15.98 GB physical, 16.48 GB virtual
____________
My stats |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 14036
ID: 53948
Credit: 475,998,733
RAC: 222,579
|
That line lists all the features the cpu has. CPUs with AVX instructions should should have AVX (or something like that) on that line.
I wonder which one represents AVX.
11/01/2012 13:05:47 | | Processor: 4 GenuineIntel Intel(R) Core(TM) i5-2500 CPU @ 3.30GHz [Family 6 Model 42 Stepping 7]
11/01/2012 13:05:47 | | Processor: 256.00 KB cache
11/01/2012 13:05:47 | | Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss htt tm pni ssse3 cx16 sse4_1 sse4_2 syscall nx lm vmx smx tm2 popcnt aes pbe
11/01/2012 13:05:47 | | OS: Microsoft Windows Server 2008 "R2": Standard x64 Edition, Service Pack 1, (06.01.7601.00)
11/01/2012 13:05:47 | | Memory: 15.98 GB physical, 16.48 GB virtual
It could be that the boinc code doesn't recognize that bit. Let me dig around and look and see what I can find.
____________
My lucky number is 75898524288+1 |
|
|
|
|
|
Another major problem seen this some times ago now with all versions tested.
CTRL+C to kill the client. Restart computer. Start client but it dont want resume from the last checkpoint. Checkpoint file is still there + workunit.
Tried this again and again. The client only get new WUs.
I have compressed this one for further investigations if a dev want to look into it.
|
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 14036
ID: 53948
Credit: 475,998,733
RAC: 222,579
|
That line lists all the features the cpu has. CPUs with AVX instructions should should have AVX (or something like that) on that line.
I wonder which one represents AVX.
11/01/2012 13:05:47 | | Processor: 4 GenuineIntel Intel(R) Core(TM) i5-2500 CPU @ 3.30GHz [Family 6 Model 42 Stepping 7]
11/01/2012 13:05:47 | | Processor: 256.00 KB cache
11/01/2012 13:05:47 | | Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss htt tm pni ssse3 cx16 sse4_1 sse4_2 syscall nx lm vmx smx tm2 popcnt aes pbe
11/01/2012 13:05:47 | | OS: Microsoft Windows Server 2008 "R2": Standard x64 Edition, Service Pack 1, (06.01.7601.00)
11/01/2012 13:05:47 | | Memory: 15.98 GB physical, 16.48 GB virtual
It could be that the boinc code doesn't recognize that bit. Let me dig around and look and see what I can find.
Yeah, at least as of client 6.12.34 (the latest stable client) it doesn't check for the AVX bit. AVX is in bit 28 of the ECX register when EAX==1 when you execute the CPUID instruction.
Here's a link to a description of all the feature bits: http://en.wikipedia.org/wiki/CPUID.
____________
My lucky number is 75898524288+1 |
|
|
HonzaVolunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1963
ID: 352
Credit: 6,398,257,295
RAC: 2,430,869
|
Yeah, at least as of client 6.12.34 (the latest stable client) it doesn't check for the AVX bit. AVX is in bit 28 of the ECX register when EAX==1 when you execute the CPUID instruction.
Yeah, it was pasted from 6.12.34. 6.10.58 and 6.13.x are working as well...since detection is in app, not BOINC client and we are using app_info.xml anyway.
____________
My stats |
|
|
rogueVolunteer developer
Send message
Joined: 8 Sep 07
Posts: 1259
ID: 12001
Credit: 18,565,548
RAC: 0
|
Another major problem seen this some times ago now with all versions tested.
CTRL+C to kill the client. Restart computer. Start client but it dont want resume from the last checkpoint. Checkpoint file is still there + workunit.
Tried this again and again. The client only get new WUs.
I have compressed this one for further investigations if a dev want to look into it.
You're the first to report this. Set debuglevel=1, reproduce, then send me the log.
BTW, what are startoption and stopoption set to? |
|
|
|
|
Another major problem seen this some times ago now with all versions tested.
CTRL+C to kill the client. Restart computer. Start client but it dont want resume from the last checkpoint. Checkpoint file is still there + workunit.
Tried this again and again. The client only get new WUs.
I have compressed this one for further investigations if a dev want to look into it.
You're the first to report this. Set debuglevel=1, reproduce, then send me the log.
BTW, what are startoption and stopoption set to?
9,9,0 is the setting. |
|
|
|
|
|
thanks for reply guys. you are awesome!!
i've got a question. as i read in wikipedia
Recent releases of GCC starting with version 4.6 (although there was a 4.3 branch with certain support) and the Intel Compiler Suite starting with version 11.1 support AVX. The Visual Studio 2010 compiler supports AVX via intrinsic and /arch:AVX switch.
does this mean that if one try to compile some app with "gcc -mavx" - the compiler will try to use more efficient instructions and this may lead to performance increase?
or one should use some asm to get a benefit from 256 bit registers?
____________
wbr, Me. Dead J. Dona
|
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 14036
ID: 53948
Credit: 475,998,733
RAC: 222,579
|
thanks for reply guys. you are awesome!!
i've got a question. as i read in wikipedia
Recent releases of GCC starting with version 4.6 (although there was a 4.3 branch with certain support) and the Intel Compiler Suite starting with version 11.1 support AVX. The Visual Studio 2010 compiler supports AVX via intrinsic and /arch:AVX switch.
does this mean that if one try to compile some app with "gcc -mavx" - the compiler will try to use more efficient instructions and this may lead to performance increase?
or one should use some asm to get a benefit from 256 bit registers?
It's possible you might see some small improvement merely by telling the compiler to use avx instructions, but the real benefit comes from designing the software to take advantage of the new instructions. You've got to write the a program that uses the new architecture. It doesn't happen automatically.
____________
My lucky number is 75898524288+1 |
|
|
|
|
|
Today, I discovered a restart problem, which only affect genefer.
In prpclient.ini, I use GFN32768 with share 100% and GFN65536 and GFN262144 as backup with share 0.
At the genefer executables line I use:
//geneferexe=genefercuda.exe
geneferexe=genefx64.exe
geneferexe=genefer.exe
geneferexe=genefer80.exe
If the backup port is used, it's better to do this work with genefx64 instead of genefer80.
What happens normally, my client connected to port GFN32768, try genefx64 > maxerror > try genefer > maxerror > try and use genefer80.
This works pretty good, all the time. So this night this host restarted, and in the morning I try to restart the work. But now with this error:
C:\prpnet\prpclient-1>prpclient.exe
[2012-01-12 06:56:04 MZ] PRPNet Client application v5.0.2 started
[2012-01-12 06:56:04 MZ] User name odicin at email address is gs@fe117.de
GenefX64 2.2.1 (x86 - 64-bit - SSE3) Copyright (C) 2001-2003, Yves Gallot
Copyright 2009-2011, Mark Rodenkirch, David Underbakke
A program for finding large probable generalized Fermat primes.
Start test of file 'work_GFN32768.in' - 06:56:04
Checkpoint saved by another build of genefer (23 19). This version cannot resta
rt with it.Genefer 2.2.1 (x86 - 32-bit - 64-b X87) Copyright (C) 2001-2003, Yves
Gallot
Copyright 2009-2011, Mark Rodenkirch, David Underbakke
A program for finding large probable generalized Fermat primes.
Start test of file 'work_GFN32768.in'.
Testing 4377322^32768+1...
Checkpoint saved by genefer x86 - 32-bit - 80-b X87. This version cannot restar
t with it
[2012-01-12 06:56:05 MZ] GFN32768: Could not find RES residue [genefer.log]. Is
genefer broken?
I wondered a while around and then I understand. The client first try genefx64, but not using the other executables afterwards. So I commented out genefx64 and genefer in prpclient.ini and try again and now the resuming of genefer80 works.
After restart, I revoked the changes in prpclient.ini and the client is running like before.
Is it possible to check the executables after a restart in the same way?
Regards Odi
____________
 |
|
|
|
|
thanks for reply guys. you are awesome!!
i've got a question. as i read in wikipedia
Recent releases of GCC starting with version 4.6 (although there was a 4.3 branch with certain support) and the Intel Compiler Suite starting with version 11.1 support AVX. The Visual Studio 2010 compiler supports AVX via intrinsic and /arch:AVX switch.
does this mean that if one try to compile some app with "gcc -mavx" - the compiler will try to use more efficient instructions and this may lead to performance increase?
or one should use some asm to get a benefit from 256 bit registers?
It's possible you might see some small improvement merely by telling the compiler to use avx instructions, but the real benefit comes from designing the software to take advantage of the new instructions. You've got to write the a program that uses the new architecture. It doesn't happen automatically.
My first naive attempt at an AVX build, just changing the compiler flags for the C compilation, produced no meaningful speed improvement. It was only when the newest gwnum library was incorporated, and other minor build-process tweaks were made, that I was able to see the reported 20-50% LLR speed-up. Others have followed with Mac and Windows builds. I'm working with others to produce a "universal" (at least for Linux) executable that should take advantage of AVX when it is available, and otherwise fall back to the existing code, so that BOINC users can avoid the app_info mess. In the meantime the app should work fine on PRPnet for 64-bit Linux (although some have had problems with AMD bulldozer CPUs).
As for telling if your cpu should support the new instructions, "sysinfo" on Linux, under the CPU tab, will show 'avx' (amongst many other things) if supported. I think CPU-Z on windows will give roughly the same information. *Very* rough guideline: if your box is more than a year old, forget it. If newer, then maybe.
--Gary
____________
"I am he as you are he as you are me and we are all together"
87*2^3496188+1 is prime! (1052460 digits)
4 is not prime! (1 digit) |
|
|
|
|
I think CPU-Z on windows will give roughly the same information.
Yes, it will, under cpu tab, "instructions".
Also, a look at intel architecture site (http://ark.intel.com) will show AVX under "Instruction Set Extensions", if it is available on a given processor. |
|
|
rroonnaallddVolunteer developer
Volunteer tester
Send message
Joined: 3 Jul 09
Posts: 1213
ID: 42893
Credit: 34,634,263
RAC: 0
|
|
Thanks for pointing to "sysinfo". I installed it in my Lubuntu32-VM and get the output directly via Xming on my Win-desktop...
____________
Best wishes. Knowledge is power. by jjwhalen
|
|
|
|
|
|
the download site seems to be down, I can't access it.... Are the files hosted anywhere else? |
|
|
pschoeferVolunteer developer
Volunteer tester
Send message
Joined: 20 Sep 05
Posts: 686
ID: 845
Credit: 3,008,700,736
RAC: 610,120
                 
|
|
I ran into a small problem using PRPNet 5.0.4 and genefercuda-boinc, while all other genefers were commented out:
[2012-02-01 08:46:14 MZ] GFN262144: Getting work from server prpnet.mine.nu at port 11002
[2012-02-01 08:46:14 MZ] GFN262144: sending [GETWORK 5.0.4 1]
[2012-02-01 08:46:14 MZ] GFN262144: sending [genefercuda-boinc]
[2012-02-01 08:46:14 MZ] GFN262144: sending [End of Message]
[2012-02-01 08:46:14 MZ] GFN262144: received [ServerInfo: 8 4.3.7 0]
[2012-02-01 08:46:18 MZ] GFN262144: received [keepalive]
[2012-02-01 08:46:21 MZ] GFN262144: received [INFO: No available candidates are left on this server.]
[2012-02-01 08:46:21 MZ] GFN262144: INFO: No available candidates are left on this server.
[2012-02-01 08:46:21 MZ] GFN262144: received [End of Message]
As soon as prpclient is allowed to use any other genefer, it's getting work and crunching it with genefercuda. Looks like the server just doesn't realize that genefercuda-boinc is a valid genefer.
____________
  |
|
|
rogueVolunteer developer
Send message
Joined: 8 Sep 07
Posts: 1259
ID: 12001
Credit: 18,565,548
RAC: 0
|
I ran into a small problem using PRPNet 5.0.4 and genefercuda-boinc, while all other genefers were commented out:
[2012-02-01 08:46:14 MZ] GFN262144: Getting work from server prpnet.mine.nu at port 11002
[2012-02-01 08:46:14 MZ] GFN262144: sending [GETWORK 5.0.4 1]
[2012-02-01 08:46:14 MZ] GFN262144: sending [genefercuda-boinc]
[2012-02-01 08:46:14 MZ] GFN262144: sending [End of Message]
[2012-02-01 08:46:14 MZ] GFN262144: received [ServerInfo: 8 4.3.7 0]
[2012-02-01 08:46:18 MZ] GFN262144: received [keepalive]
[2012-02-01 08:46:21 MZ] GFN262144: received [INFO: No available candidates are left on this server.]
[2012-02-01 08:46:21 MZ] GFN262144: INFO: No available candidates are left on this server.
[2012-02-01 08:46:21 MZ] GFN262144: received [End of Message]
As soon as prpclient is allowed to use any other genefer, it's getting work and crunching it with genefercuda. Looks like the server just doesn't realize that genefercuda-boinc is a valid genefer.
Something isn't quite right. The server looks for the string "genefercuda", not "genefercuda-boinc". Does that need to be changed on the server? |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 14036
ID: 53948
Credit: 475,998,733
RAC: 222,579
|
|
Have you tried renaming GeneferCUDA-boinc-windows to just GeneferCUDA, and NOT changing the prpclient.ini file? That's the way I do it.
____________
My lucky number is 75898524288+1 |
|
|
pschoeferVolunteer developer
Volunteer tester
Send message
Joined: 20 Sep 05
Posts: 686
ID: 845
Credit: 3,008,700,736
RAC: 610,120
|
|
rogue wrote: Something isn't quite right. The server looks for the string "genefercuda", not "genefercuda-boinc". Does that need to be changed on the server?
Michael's latest version reports as "GeneferCUDA-boinc 1.06 (CUDA 3.2) based on GeneferCUDA 1.049 and Genefer 2.2.1". The older version I was using before reported as "GeneferCUDA 0.97 based on Genefer v2.2.0". I guess only the bold part of that string is reported to the server.
So if the server is expected to support a client with only Michael's GeneferCUDA specified in the prpclient.ini, either the server needs to be changed or the client has to cut off the '-boinc' when reporting the available executables.
Michael Goetz wrote: Have you tried renaming GeneferCUDA-boinc-windows to just GeneferCUDA, and NOT changing the prpclient.ini file? That's the way I do it.
That's what I did, too. Does it work for you if no other version of genefer is specified in the prpclient.ini?
____________
|
|
|
Crun-chiVolunteer tester
Send message
Joined: 25 Nov 09
Posts: 3245
ID: 50683
Credit: 152,646,050
RAC: 18,212
           
|
|
Great I am not insane :)
With Genefer Boinc Cuda 1.06 you cannot get any work, but you can get WU with clasic "PRPNet" GeneferCuda :)
____________
92*10^1585996-1 NEAR-REPDIGIT PRIME :) :) :)
4 * 650^498101-1 CRUS PRIME
2022202116^131072+1 GENERALIZED FERMAT
Proud member of team Aggie The Pew. Go Aggie! |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 14036
ID: 53948
Credit: 475,998,733
RAC: 222,579
|
|
This morning, while the BOINC server was out of GFN WUs, I used the new boinc 1.06 Genefer app on PRPNET and it definitely works just fine. I had previously been using 1.05, so that works as well.
This is with PRPNet 5.0.4. Earlier versions of the client might not work with the new GeneferCUDA.
You need to rename the executable to GeneferCUDA.exe, and you need to have the two CUDA 3.2 dlls in the directory, but that should be the only thing you need to do. It's basically plug and play.
____________
My lucky number is 75898524288+1 |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 14036
ID: 53948
Credit: 475,998,733
RAC: 222,579
|
Michael Goetz wrote: Have you tried renaming GeneferCUDA-boinc-windows to just GeneferCUDA, and NOT changing the prpclient.ini file? That's the way I do it.
That's what I did, too. Does it work for you if no other version of genefer is specified in the prpclient.ini?
I don't know. I have all 4 Genefer lines in prpclient.ini uncommented.
geneferexe=genefercuda.exe
geneferexe=genefx64.exe
geneferexe=genefer.exe
geneferexe=genefer80.exe
Genefercuda.exe, in my case, is actually the boinc version.
If you have the other lines commented out, and genefercuda can't run for some reason (e.g., the dll's are missing), perhaps you might see the behavior you're describing.
____________
My lucky number is 75898524288+1 |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 14036
ID: 53948
Credit: 475,998,733
RAC: 222,579
|
|
Mark,
Looking at my logs, I noticed something odd. This has nothing to do with the problem of not getting work.
I stopped and restarted this WU just before it finished (apparantly 13 seconds before it finished.) This is from prpclient.log:
[2012-02-01 11:18:33 EST] PRPNet Client application v5.0.4 started
[2012-02-01 11:18:33 EST] User name Michael_Goetz at email address is hidden
[2012-02-01 11:18:46 EST] GFN262144: 452582^262144+1 is not prime. Residue 41c513d1b5107022
[2012-02-01 11:18:46 EST] Total Time: 0:00:13 Total Work Units: 1 Special Results Found: 0
[2012-02-01 11:18:46 EST] GFN262144: Returning work to server prpnet.mine.nu at port 11002
[2012-02-01 11:18:47 EST] GFN262144: INFO: Test for 452582^262144+1 was accepted
[2012-02-01 11:18:47 EST] GFN262144: INFO: All 1 test results were accepted
[2012-02-01 11:18:47 EST] Client shutdown complete
This is the corresponding entry from test_results.log:
[2012-02-01 11:18:46 EST] Server: GFN262144, Candidate: 452582^262144+1 Program: genefercuda.exe Residue: 41c513d1b5107022 Time: 5183 seconds
Is that the expected behavior?
____________
My lucky number is 75898524288+1 |
|
|
rogueVolunteer developer
Send message
Joined: 8 Sep 07
Posts: 1259
ID: 12001
Credit: 18,565,548
RAC: 0
|
Mark,
Looking at my logs, I noticed something odd. This has nothing to do with the problem of not getting work.
I stopped and restarted this WU just before it finished (apparantly 13 seconds before it finished.) This is from prpclient.log:
[2012-02-01 11:18:33 EST] PRPNet Client application v5.0.4 started
[2012-02-01 11:18:33 EST] User name Michael_Goetz at email address is hidden
[2012-02-01 11:18:46 EST] GFN262144: 452582^262144+1 is not prime. Residue 41c513d1b5107022
[2012-02-01 11:18:46 EST] Total Time: 0:00:13 Total Work Units: 1 Special Results Found: 0
[2012-02-01 11:18:46 EST] GFN262144: Returning work to server prpnet.mine.nu at port 11002
[2012-02-01 11:18:47 EST] GFN262144: INFO: Test for 452582^262144+1 was accepted
[2012-02-01 11:18:47 EST] GFN262144: INFO: All 1 test results were accepted
[2012-02-01 11:18:47 EST] Client shutdown complete
This is the corresponding entry from test_results.log:
[2012-02-01 11:18:46 EST] Server: GFN262144, Candidate: 452582^262144+1 Program: genefercuda.exe Residue: 41c513d1b5107022 Time: 5183 seconds
Is that the expected behavior?
Yes. The 0:00:13 is clock time for the client, not clock time for any PRP tests done by the client. |
|
|
rogueVolunteer developer
Send message
Joined: 8 Sep 07
Posts: 1259
ID: 12001
Credit: 18,565,548
RAC: 0
|
This morning, while the BOINC server was out of GFN WUs, I used the new boinc 1.06 Genefer app on PRPNET and it definitely works just fine. I had previously been using 1.05, so that works as well.
This is with PRPNet 5.0.4. Earlier versions of the client might not work with the new GeneferCUDA.
You need to rename the executable to GeneferCUDA.exe, and you need to have the two CUDA 3.2 dlls in the directory, but that should be the only thing you need to do. It's basically plug and play.
What the client sends to the server is independent of the executable file name. It pulls the name from the executable when ti runs it. I'm curious why yours works and his doesn't. If you turn on logging, what does your client send to the server? |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 14036
ID: 53948
Credit: 475,998,733
RAC: 222,579
|
|
hold on. let me turn on logging and fire up prppclient....
AHAH! I understand what's happening:
[2012-02-01 12:29:52 EST] PRPNet Client application v5.0.4 started
[2012-02-01 12:29:52 EST] User name Michael_Goetz at email address is mike@asgoo
dasitgoetz.com
[2012-02-01 12:29:52 EST] GFN262144: sending [FROM mike@asgoodasitgoetz.com Vipe
r-CUDA-1 Michael_Goetz 5.0.4 SETI.USA]
[2012-02-01 12:29:52 EST] GFN262144: received [Connected to server 4.3.7]
[2012-02-01 12:29:52 EST] GFN262144: Getting work from server prpnet.mine.nu at
port 11002
[2012-02-01 12:29:52 EST] GFN262144: sending [GETWORK 5.0.4 1]
[2012-02-01 12:29:52 EST] GFN262144: sending [llr]
[2012-02-01 12:29:52 EST] GFN262144: sending [pfgw]
[2012-02-01 12:29:52 EST] GFN262144: sending [genefercuda-boinc]
[2012-02-01 12:29:52 EST] GFN262144: sending [genefx64]
[2012-02-01 12:29:52 EST] GFN262144: sending [genefer]
[2012-02-01 12:29:52 EST] GFN262144: sending [genefer80]
[2012-02-01 12:29:52 EST] GFN262144: sending [wwww]
[2012-02-01 12:29:52 EST] GFN262144: sending [End of Message]
[2012-02-01 12:29:53 EST] GFN262144: received [ServerInfo: 8 4.3.7 0]
[2012-02-01 12:29:53 EST] GFN262144: received [WorkUnit: 453196^262144+1 1328117
528 453196 262144]
[2012-02-01 12:29:53 EST] GFN262144: sending [Received: 453196^262144+1]
[2012-02-01 12:29:54 EST] GFN262144: received [End of Message]
[2012-02-01 12:29:54 EST] GFN262144: PRPNet server is version 4.3.7
[2012-02-01 12:29:54 EST] GFN262144: received [Start Greeting]
[2012-02-01 12:29:54 EST] GFN262144: received [Hi! Welcome to PrimeGrid's GFN 26
2144 Prime Search.]
Hi! Welcome to PrimeGrid's GFN 262144 Prime Search.
[2012-02-01 12:29:54 EST] GFN262144: received [End Greeting]
[2012-02-01 12:29:54 EST] GFN262144: closing socket
GeneferCUDA-boinc 1.06 (CUDA3.2) based on GeneferCUDA 1.049 and Genefer 2.2.1
Copyright (C) 2001-2003, Yves Gallot (v1.3)
Copyright (C) 2009-2011, Mark Rodenkirch, David Underbakke (v2.2.1)
Copyright (C) 2010-2012, Shoichiro Yamada (CUDA)
Portions of this software written by Michael Goetz 2011-2012 (BOINC)
A program for finding large probable generalized Fermat primes.
Command line: genefercuda.exe work_GFN262144.in
Start test of file 'work_GFN262144.in' - 12:29:54
Testing 453196^262144+1...
maxErr during b^N initialization = 0.0000 (0.237 seconds).
Testing 453196^262144+1... 4915200 steps to go
The server doesn't recognize GeneferCUDA-boinc -- but it does recognize the other Genefer programs.
So it will send work as long as it sees one of the Genefers it DOES recognize, and the boinc version of GeneferCUDA will execute it, but if ONLY geneferCUDA is there the server doesn't see anything it knows, so it doesn't send any work.
The workaround should be to uncomment any of the other Genefer programs, and this explains why it has always worked for me but isn't working for some others.
____________
My lucky number is 75898524288+1 |
|
|
Crun-chiVolunteer tester
Send message
Joined: 25 Nov 09
Posts: 3245
ID: 50683
Credit: 152,646,050
RAC: 18,212
|
The workaround should be to uncomment any of the other Genefer programs, and this explains why it has always worked for me but isn't working for some others.
Yes, in this case it works. If only GeneferCUDA is in directory it doesnot works!
____________
92*10^1585996-1 NEAR-REPDIGIT PRIME :) :) :)
4 * 650^498101-1 CRUS PRIME
2022202116^131072+1 GENERALIZED FERMAT
Proud member of team Aggie The Pew. Go Aggie! |
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
The workaround should be to uncomment any of the other Genefer programs, and this explains why it has always worked for me but isn't working for some others.
Also, a non-BOINC build of GeneferCUDA would easily solve this. ;)
____________
|
|
|
Crun-chiVolunteer tester
Send message
Joined: 25 Nov 09
Posts: 3245
ID: 50683
Credit: 152,646,050
RAC: 18,212
|
|
I agree with you, but I got around 500 seconds faster run time with 1.06 , so it will be good to make new faster build with block switch :)
____________
92*10^1585996-1 NEAR-REPDIGIT PRIME :) :) :)
4 * 650^498101-1 CRUS PRIME
2022202116^131072+1 GENERALIZED FERMAT
Proud member of team Aggie The Pew. Go Aggie! |
|
|
rogueVolunteer developer
Send message
Joined: 8 Sep 07
Posts: 1259
ID: 12001
Credit: 18,565,548
RAC: 0
|
|
I have fixed this in the client because "genefercuda" and "genefercuda-boinc" are really the same program. You can d/l 5.0.5 from here:
http://home.roadrunner.com/~mrodenkirch/prpnet_5.0.5.zip
The zip file has a Windows build, but not a Mac/Linux build. If you need one, then you have to wait for John. |
|
|
pschoeferVolunteer developer
Volunteer tester
Send message
Joined: 20 Sep 05
Posts: 686
ID: 845
Credit: 3,008,700,736
RAC: 610,120
|
I have fixed this in the client because "genefercuda" and "genefercuda-boinc" are really the same program. You can d/l 5.0.5 from here:
Thanks. It's working now.
____________
|
|
|
samuel7Volunteer tester
Send message
Joined: 1 May 09
Posts: 89
ID: 39425
Credit: 257,425,010
RAC: 0
       
|
|
The v5.0.5 Windows-GPU archive at http://uwin.mine.nu/PRPNet/ includes CUDA 4.0 dll's. To run any of Michael's versions of GeneferCUDA you need the 3.2 dll's.
____________
 |
|
|
rroonnaallddVolunteer developer
Volunteer tester
Send message
Joined: 3 Jul 09
Posts: 1213
ID: 42893
Credit: 34,634,263
RAC: 0
|
The v5.0.5 Windows-GPU archive at http://uwin.mine.nu/PRPNet/ includes CUDA 4.0 dll's. To run any of Michael's versions of GeneferCUDA you need the 3.2 dll's.
Copy and renaming both files to the needed version should solve this "problem"...
____________
Best wishes. Knowledge is power. by jjwhalen
|
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 14036
ID: 53948
Credit: 475,998,733
RAC: 222,579
|
The v5.0.5 Windows-GPU archive at http://uwin.mine.nu/PRPNet/ includes CUDA 4.0 dll's. To run any of Michael's versions of GeneferCUDA you need the 3.2 dll's.
Copy and renaming both files to the needed version should solve this "problem"...
To clarify,
What you should do is download the correct DLL's. They can be downloaded from http://www.primegrid.com/download and place those fies in the same directory from which you are running GeneferCUDA.
You can merely rename the 4.0 dlls to the names of the 3.2 dlls -- that would indeed work -- but doing so would negate the purpose of using the 3.2 dlls: The 3.2 dlls are about 5 to 10 percent faster than the 4.0 dlls.
____________
My lucky number is 75898524288+1 |
|
|
samuel7Volunteer tester
Send message
Joined: 1 May 09
Posts: 89
ID: 39425
Credit: 257,425,010
RAC: 0
|
The v5.0.5 Windows-GPU archive at http://uwin.mine.nu/PRPNet/ includes CUDA 4.0 dll's. To run any of Michael's versions of GeneferCUDA you need the 3.2 dll's.
Copy and renaming both files to the needed version should solve this "problem"...
To clarify,
What you should do is download the correct DLL's. They can be downloaded from http://www.primegrid.com/download and place those fies in the same directory from which you are running GeneferCUDA.
You can merely rename the 4.0 dlls to the names of the 3.2 dlls -- that would indeed work -- but doing so would negate the purpose of using the 3.2 dlls: The 3.2 dlls are about 5 to 10 percent faster than the 4.0 dlls.
As a test I tried renaming 4.0 --> 3.2 but then it complained that the 4.0 dlls were missing! There's no problem running without 4.0 if the 3.2 files are present.
P.S. Copying the correct dlls to the working directory was clear to me, but it's good to have the info on the board for other users. My message was just a hint to John to update the archive.
____________
|
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
The v5.0.5 Windows-GPU archive at http://uwin.mine.nu/PRPNet/ includes CUDA 4.0 dll's. To run any of Michael's versions of GeneferCUDA you need the 3.2 dll's.
Windows PRPClient packages update to the 3.2 dll's.
____________
|
|
|
samuel7Volunteer tester
Send message
Joined: 1 May 09
Posts: 89
ID: 39425
Credit: 257,425,010
RAC: 0
|
|
Is it possible to upgrade the prpnet ports to server version 5.0.2+? Instances like this would be avoided:
[2012-02-06 11:32:27 Sn] GFN524288: Getting work from server prpnet.mine.nu at port 11001
[2012-02-06 11:32:46 Sn] GFN524288: INFO: No available candidates are left on this server.
[2012-02-06 11:32:46 Sn] GFN524288: PRPNet server is version 4.3.7
[2012-02-06 11:32:46 Sn] GFN524288: nothing was received on socket after 1 seconds
[2012-02-06 14:10:32 Sn] GFN524288: 101848^524288+1 is not prime. Residue b97a02094e5d815a
[2012-02-06 14:10:32 Sn] Total Time: 5:16:03 Total Work Units: 2 Special Results Found: 0
[2012-02-06 14:10:33 Sn] GFN524288: Returning work to server prpnet.mine.nu at port 11001
[2012-02-06 14:10:33 Sn] GFN524288: INFO: Test for 101848^524288+1 was ignored. Candidate and/or test was not found
[2012-02-06 14:10:34 Sn] GFN524288: INFO: 0 of 1 test results were accepted
Well, at least this candidate can be marked as double checked when the other test is returned :-/
____________
|
|
|
|
|
|
Newbie to prpnet here so I have a quick question.
Is there a way to see what tasks I have completed and if they are getting credited to me and my team?
____________
 |
|
|
rogueVolunteer developer
Send message
Joined: 8 Sep 07
Posts: 1259
ID: 12001
Credit: 18,565,548
RAC: 0
|
Newbie to prpnet here so I have a quick question.
Is there a way to see what tasks I have completed and if they are getting credited to me and my team?
To see specific tasks you can look at the completed_tests.log file on your local machine. As for where they are credited you can look at the user_stats.html page, which basically this:
http://server:port/user_stats.html
in your browser. |
|
|
|
|
Newbie to prpnet here so I have a quick question.
Is there a way to see what tasks I have completed and if they are getting credited to me and my team?
To see specific tasks you can look at the completed_tests.log file on your local machine. As for where they are credited you can look at the user_stats.html page, which basically this:
http://server:port/user_stats.html
in your browser.
Great, thanks. One other question. When I check the server stats page like this: http://prpnet.primegrid.com:12003/server_stats.html it says 9 days before the server runs out of work but it won't send me any from this port.
I have this configured in my config file:
server=GFN65536:25:1:prpnet.primegrid.com:12003
____________
|
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 14036
ID: 53948
Credit: 475,998,733
RAC: 222,579
|
One other question. When I check the server stats page like this: http://prpnet.primegrid.com:12003/server_stats.html it says 9 days before the server runs out of work but it won't send me any from this port.
I have this configured in my config file:
server=GFN65536:25:1:prpnet.primegrid.com:12003
You can only crunch 65536 on your CPU. The current b value is too high for GeneferCUDA. The server *might* be smart enough to know that and perhaps it's not sending you any work if you only have GeneferCUDA enabled, but not any of the programs that can handle that large b.
____________
My lucky number is 75898524288+1 |
|
|
|
|
One other question. When I check the server stats page like this: http://prpnet.primegrid.com:12003/server_stats.html it says 9 days before the server runs out of work but it won't send me any from this port.
I have this configured in my config file:
server=GFN65536:25:1:prpnet.primegrid.com:12003
You can only crunch 65536 on your CPU. The current b value is too high for GeneferCUDA. The server *might* be smart enough to know that and perhaps it's not sending you any work if you only have GeneferCUDA enabled, but not any of the programs that can handle that large b.
This one should work though correct: server=GFN32768:25:1:prpnet.primegrid.com:12005
____________
|
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 14036
ID: 53948
Credit: 475,998,733
RAC: 222,579
|
One other question. When I check the server stats page like this: http://prpnet.primegrid.com:12003/server_stats.html it says 9 days before the server runs out of work but it won't send me any from this port.
I have this configured in my config file:
server=GFN65536:25:1:prpnet.primegrid.com:12003
You can only crunch 65536 on your CPU. The current b value is too high for GeneferCUDA. The server *might* be smart enough to know that and perhaps it's not sending you any work if you only have GeneferCUDA enabled, but not any of the programs that can handle that large b.
This one should work though correct: server=GFN32768:25:1:prpnet.primegrid.com:12005
Nope. That one's even worse. GeneferCUDA can currently only work on 262144 or 524288 (or higher if they open up ports at higher N). The 32K and 64K ports are at a b level that's too high for the CUDA app.
____________
My lucky number is 75898524288+1 |
|
|
|
|
|
ok well these are the servers I now have defined. I only get work from this one:
server=121:0:1:prpnet.primegrid.com:12001
Nothing from these:
server=PPSEhigh:0:1:prpnet.primegrid.com:12007
server=27:0:1:prpnet.primegrid.com:12006
server=GFN262144:50:1:prpnet.mine.nu:11002
server=GFN524288:50:1:prpnet.mine.nu:11001
I have this configured:
llrexe=llrCUDA.exe
pfgwexe=pfgw64.exe
geneferexe=geneferCUDA-boinc-windows
anything wrong with this so far?
____________
|
|
|
|
|
|
The line
geneferexe=geneferCUDA-boinc-windows
should end in .exe
Also, PPSEhigh has been completed. |
|
|
|
|
The line
geneferexe=geneferCUDA-boinc-windows
should end in .exe
Also, PPSEhigh has been completed.
Fixed, TYVM!
____________
|
|
|
|
|
|
Ok, I have been running and completing 121 port WUs, using llrCUDA but now that I got the right genefer ports I selected that but it won't run I get this:
Command line: geneferCUDA-boinc-windows.exe -d 0 work_GFN262144.in
Start test of file '-d' - 01:37:37
Testing 469818^262144+1...
maxErr exceeded during initialization at 262144, 0.5000 > 0.4500
[2012-02-17 01:37:37 PST] GFN262144: No data in file [genefer.log]. Is genefer
broken?
____________
|
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 14036
ID: 53948
Credit: 475,998,733
RAC: 222,579
|
Ok, I have been running and completing 121 port WUs, using llrCUDA but now that I got the right genefer ports I selected that but it won't run I get this:
Command line: geneferCUDA-boinc-windows.exe -d 0 work_GFN262144.in
Start test of file '-d' - 01:37:37
Testing 469818^262144+1...
maxErr exceeded during initialization at 262144, 0.5000 > 0.4500
[2012-02-17 01:37:37 PST] GFN262144: No data in file [genefer.log]. Is genefer
broken?
It looks like you're having the same problem here as on the Boinc side.
____________
My lucky number is 75898524288+1 |
|
|
|
|
Ok, I have been running and completing 121 port WUs, using llrCUDA but now that I got the right genefer ports I selected that but it won't run I get this:
Command line: geneferCUDA-boinc-windows.exe -d 0 work_GFN262144.in
Start test of file '-d' - 01:37:37
Testing 469818^262144+1...
maxErr exceeded during initialization at 262144, 0.5000 > 0.4500
[2012-02-17 01:37:37 PST] GFN262144: No data in file [genefer.log]. Is genefer
broken?
It looks like you're having the same problem here as on the Boinc side.
Ok, that is what I figured, so I will just run the llrCUDA 121 wus until I work that out.
____________
|
|
|
|
|
|
How can I tell the 5.x client do not get new work while server setting is at 0% only? Its more than confused vs. 4.x client. Perhaps its better with a new setting = get new work on/off. |
|
|
rogueVolunteer developer
Send message
Joined: 8 Sep 07
Posts: 1259
ID: 12001
Credit: 18,565,548
RAC: 0
|
How can I tell the 5.x client do not get new work while server setting is at 0% only? Its more than confused vs. 4.x client. Perhaps its better with a new setting = get new work on/off.
You need to comment out the line to prevent it from getting new work from the server. At 0%, it can get work if unable to connect to servers with pct > 0.
If that isn't the behavior you are seeing, I suggest setting debuglevel=3 to see why it is getting work from that server. |
|
|
|
|
How can I tell the 5.x client do not get new work while server setting is at 0% only? Its more than confused vs. 4.x client. Perhaps its better with a new setting = get new work on/off.
You need to comment out the line to prevent it from getting new work from the server. At 0%, it can get work if unable to connect to servers with pct > 0.
If that isn't the behavior you are seeing, I suggest setting debuglevel=3 to see why it is getting work from that server.
The comment out is only working if the client is closed. If its running it ignores the setting.
Compiled a 64bit v5.0.6
[2012-02-22 18:26:53 MZ] PRPNet Client application v5.0.6 started
[2012-02-22 18:26:53 MZ] User name rebirther at email address is xxx[2012-02-22 18:26:53 MZ] PPSElow: Returning work. currentworkunits=0, completedworkunits=0, quitOption=3
[2012-02-22 18:26:53 MZ] in FindAnyIncompleteWork
[2012-02-22 18:26:53 MZ] Work: PPSElow, HaveInProgressTest = false, CanDoAnotherTest = false
[2012-02-22 18:26:53 MZ] in FindNextServerForWork: total time for client=0 seconds
[2012-02-22 18:26:53 MZ] Work: PPSElow, no work done yet, target pct work done=100
[2012-02-22 18:26:53 MZ] PPSElow: Getting work from server prpnet.mine.nu at port 12000
[2012-02-22 18:26:57 MZ] PPSElow: AddWorkUnitToList for 8853*2^323596+1
[2012-02-22 18:26:57 MZ] PPSElow: IsWorkUnitCompleted for 8853*2^323596+1. Returning false (not completed, main)
[2012-02-22 18:26:58 MZ] PPSElow: PRPNet server is version 4.3.6
[2012-02-22 18:26:58 MZ] PPSElow: AddWorkUnitToList for 8853*2^323596+1
[2012-02-22 18:26:58 MZ] PPSElow: IsWorkUnitCompleted for 8853*2^323596+1. Returning false (not completed, main)
[2012-02-22 18:26:58 MZ] PPSElow: IsWorkUnitCompleted for 8853*2^323596+1. Returning false (not completed, main)
[2012-02-22 18:26:58 MZ] PPSElow: CanDoAnotherTest for 8853*2^323596+1. Returning true (not done)
[2012-02-22 18:26:58 MZ] PPSElow: GetNextIncompleteWorkUnit looking for work. In Progress=false
[2012-02-22 18:26:58 MZ] PPSElow: IsWorkUnitCompleted for 8853*2^323596+1. Returning false (not completed, main)
[2012-02-22 18:26:58 MZ] PPSElow: GetNextIncompleteWorkUnit found 8853*2^323596+1
[2012-02-22 18:27:47 MZ] 8853*2^323596+1 is composite, a primality test is not needed
[2012-02-22 18:27:47 MZ] 8853*2^323596+1 is not PRP or prime. GFN divisibility check skipped
[2012-02-22 18:27:47 MZ] PPSElow: 8853*2^323596+1 is not prime. Residue 8A2C6F841F870ED6
[2012-02-22 18:27:47 MZ] PPSElow: ProcessWorkUnit completed testing 8853*2^323596+1
[2012-02-22 18:27:47 MZ] PPSElow: AddWorkUnitToList for 8853*2^323596+1
[2012-02-22 18:27:47 MZ] PPSElow: IsWorkUnitCompleted for 8853*2^323596+1. Returning true (completed)
[2012-02-22 18:27:47 MZ] PPSElow: IsWorkUnitCompleted for 8853*2^323596+1. Returning true (completed)
[2012-02-22 18:27:47 MZ] PPSElow: CanDoAnotherTest. Returning false (all done)
[2012-02-22 18:27:47 MZ] Total Time: 0:00:55 Total Work Units: 1 Special Results Found: 0
[2012-02-22 18:27:47 MZ] PPSElow: Returning work. currentworkunits=1, completedworkunits=1, quitOption=1
[2012-02-22 18:27:47 MZ] PPSElow: Returning work to server prpnet.mine.nu at port 12000
[2012-02-22 18:27:47 MZ] PPSElow: IsWorkUnitCompleted for 8853*2^323596+1. Returning true (completed)
[2012-02-22 18:27:47 MZ] PPSElow: INFO: Test for 8853*2^323596+1 was accepted
[2012-02-22 18:27:47 MZ] PPSElow: DeleteWorkUnit for 8853*2^323596+1
[2012-02-22 18:27:47 MZ] PPSElow: INFO: All 1 test results were accepted
[2012-02-22 18:27:47 MZ] in FindAnyIncompleteWork
[2012-02-22 18:27:47 MZ] PPSElow: CanDoAnotherTest. Returning false (none in progress)
[2012-02-22 18:27:47 MZ] PPSElow: CanDoAnotherTest. Returning false (all done)
[2012-02-22 18:27:47 MZ] Work: PPSElow, HaveInProgressTest = false, CanDoAnotherTest = false
[2012-02-22 18:27:47 MZ] in FindNextServerForWork: total time for client=49 seconds
[2012-02-22 18:27:47 MZ] Work: PPSElow, time=49, pct time=99.187755, work pct=100
[2012-02-22 18:27:48 MZ] PPSElow: Getting work from server prpnet.mine.nu at port 12000
[2012-02-22 18:27:52 MZ] PPSElow: AddWorkUnitToList for 2227*2^323598+1
[2012-02-22 18:27:52 MZ] PPSElow: IsWorkUnitCompleted for 2227*2^323598+1. Returning false (not completed, main)
[2012-02-22 18:27:53 MZ] PPSElow: PRPNet server is version 4.3.6
[2012-02-22 18:27:53 MZ] PPSElow: AddWorkUnitToList for 2227*2^323598+1
[2012-02-22 18:27:53 MZ] PPSElow: IsWorkUnitCompleted for 2227*2^323598+1. Returning false (not completed, main)
[2012-02-22 18:27:53 MZ] PPSElow: IsWorkUnitCompleted for 2227*2^323598+1. Returning false (not completed, main)
[2012-02-22 18:27:53 MZ] PPSElow: CanDoAnotherTest for 2227*2^323598+1. Returning true (not done)
[2012-02-22 18:27:53 MZ] PPSElow: GetNextIncompleteWorkUnit looking for work. In Progress=false
[2012-02-22 18:27:53 MZ] PPSElow: IsWorkUnitCompleted for 2227*2^323598+1. Returning false (not completed, main)
[2012-02-22 18:27:53 MZ] PPSElow: GetNextIncompleteWorkUnit found 2227*2^323598+1
[2012-02-22 18:28:43 MZ] 2227*2^323598+1 is composite, a primality test is not needed
[2012-02-22 18:28:43 MZ] 2227*2^323598+1 is not PRP or prime. GFN divisibility check skipped
[2012-02-22 18:28:43 MZ] PPSElow: 2227*2^323598+1 is not prime. Residue FFB28B075D568044
[2012-02-22 18:28:43 MZ] PPSElow: ProcessWorkUnit completed testing 2227*2^323598+1
[2012-02-22 18:28:43 MZ] PPSElow: AddWorkUnitToList for 2227*2^323598+1
[2012-02-22 18:28:43 MZ] PPSElow: IsWorkUnitCompleted for 2227*2^323598+1. Returning true (completed)
[2012-02-22 18:28:43 MZ] PPSElow: IsWorkUnitCompleted for 2227*2^323598+1. Returning true (completed)
[2012-02-22 18:28:43 MZ] PPSElow: CanDoAnotherTest. Returning false (all done)
[2012-02-22 18:28:43 MZ] Total Time: 0:01:51 Total Work Units: 2 Special Results Found: 0
[2012-02-22 18:28:43 MZ] PPSElow: Returning work. currentworkunits=1, completedworkunits=1, quitOption=1
[2012-02-22 18:28:43 MZ] PPSElow: Returning work to server prpnet.mine.nu at port 12000
[2012-02-22 18:28:43 MZ] PPSElow: IsWorkUnitCompleted for 2227*2^323598+1. Returning true (completed)
[2012-02-22 18:28:43 MZ] PPSElow: INFO: Test for 2227*2^323598+1 was accepted
[2012-02-22 18:28:43 MZ] PPSElow: DeleteWorkUnit for 2227*2^323598+1
[2012-02-22 18:28:43 MZ] PPSElow: INFO: All 1 test results were accepted
[2012-02-22 18:28:43 MZ] Due to changes int the prpclient.ini file, work classes have been reloaded
[2012-02-22 18:28:43 MZ] in FindAnyIncompleteWork
[2012-02-22 18:28:43 MZ] Work: PPSElow, HaveInProgressTest = false, CanDoAnotherTest = false
[2012-02-22 18:28:43 MZ] in FindNextServerForWork: total time for client=99 seconds
[2012-02-22 18:28:43 MZ] Work: PPSElow, time=0, pct time=0.000000, work pct=100
|
|
|
rogueVolunteer developer
Send message
Joined: 8 Sep 07
Posts: 1259
ID: 12001
Credit: 18,565,548
RAC: 0
|
|
I guess I'm not understanding the issue you are seeing. Let's continue this via e-mail (not PM)
What I will need is the following:
1) A description of the issue, i.e. what you think is supposed to happen vs. what is actually happening.
2) The relevant lines from prpclient.ini, i.e. server=
3) debulevel=3 in prpclient.ini
4) A demonstration of the issue in the log |
|
|
|
|
|
Are the WUs from prpnet doublechecked?
If some host uploading invalid results, than we can miss prime? |
|
|
|
|
|
I wondered about the hidden primes: http://u-g-f.de/PRPNet/hidden_primes.php. Normally there tested on primality after the probable found. Depends this on the last changes in the prpnet client?
Since primality tests take a long time for primorials/factorials, and as the pfgw doesn't checkpoint primality tests, the client will no longer do them for numbers of those forms.
I think at these both ports the lengths of the primes are not so long that a primality test is critical.
Regards Odi
____________
|
|
|
rogueVolunteer developer
Send message
Joined: 8 Sep 07
Posts: 1259
ID: 12001
Credit: 18,565,548
RAC: 0
|
I wondered about the hidden primes: http://u-g-f.de/PRPNet/hidden_primes.php. Normally there tested on primality after the probable found. Depends this on the last changes in the prpnet client?
Since primality tests take a long time for primorials/factorials, and as the pfgw doesn't checkpoint primality tests, the client will no longer do them for numbers of those forms.
I think at these both ports the lengths of the primes are not so long that a primality test is critical.
Regards Odi
Hidden means that it was reported to the server, but the server won't list it yet. This give the user who found the prime the opportunity to make their own announcement and to add to the Prime Pages. |
|
|
|
|
|
I know... but curiously, there hidden since a couple of days, maybe a week. Depending on my experience on these ports, they unhide if the primarily test with pfgw is done.
Regards Odi
____________
|
|
|
Scott BrownVolunteer moderator
Project administrator
Volunteer tester
Project scientist
Send message
Joined: 17 Oct 05
Posts: 2416
ID: 1178
Credit: 19,978,285,056
RAC: 19,305,139
|
I wondered about the hidden primes: http://u-g-f.de/PRPNet/hidden_primes.php. Normally there tested on primality after the probable found. Depends this on the last changes in the prpnet client?
Since primality tests take a long time for primorials/factorials, and as the pfgw doesn't checkpoint primality tests, the client will no longer do them for numbers of those forms.
I think at these both ports the lengths of the primes are not so long that a primality test is critical.
Regards Odi
Hidden means that it was reported to the server, but the server won't list it yet. This give the user who found the prime the opportunity to make their own announcement and to add to the Prime Pages.
Not exactly this time. For the GFN65536 ones this the case (no notification e-mails yet for me at least). However, the GFN32768 primes have been reported and are listed now on the top5000 list (one of them for quite some time now). I also am wondering why they are still hidden?
____________
141941*2^4299438-1 is prime!
|
|
|
rogueVolunteer developer
Send message
Joined: 8 Sep 07
Posts: 1259
ID: 12001
Credit: 18,565,548
RAC: 0
|
Not exactly this time. For the GFN65536 ones this the case (no notification e-mails yet for me at least). However, the GFN32768 primes have been reported and are listed now on the top5000 list (one of them for quite some time now). I also am wondering why they are still hidden?
The server doesn't know when they are reported. The length of time to "hide" a prime is configured on the server. It is possible that the configured time is too long. That is something for Lennart to decide. |
|
|
Scott BrownVolunteer moderator
Project administrator
Volunteer tester
Project scientist
Send message
Joined: 17 Oct 05
Posts: 2416
ID: 1178
Credit: 19,978,285,056
RAC: 19,305,139
|
The server doesn't know when they are reported. The length of time to "hide" a prime is configured on the server. It is possible that the configured time is too long. That is something for Lennart to decide.
Well, then maybe something hung up on the server, because the GFN32768's I found on very different days (March 7th and March 15th) both moved from hidden status today.
____________
141941*2^4299438-1 is prime!
|
|
|
rogueVolunteer developer
Send message
Joined: 8 Sep 07
Posts: 1259
ID: 12001
Credit: 18,565,548
RAC: 0
|
The server doesn't know when they are reported. The length of time to "hide" a prime is configured on the server. It is possible that the configured time is too long. That is something for Lennart to decide.
Well, then maybe something hung up on the server, because the GFN32768's I found on very different days (March 7th and March 15th) both moved from hidden status today.
I have two possible explanations. First, the page was cached by your browser (which is not likely). Second, Lennart did some of his magic and changed the settings on the serve.r |
|
|
|
|
The server doesn't know when they are reported. The length of time to "hide" a prime is configured on the server. It is possible that the configured time is too long. That is something for Lennart to decide.
Well, then maybe something hung up on the server, because the GFN32768's I found on very different days (March 7th and March 15th) both moved from hidden status today.
I have two possible explanations. First, the page was cached by your browser (which is not likely). Second, Lennart did some of his magic and changed the settings on the serve.r
I was there and changed them to be shown :)
Lennart |
|
|
|
|
|
Can I ask why this port is set at 4 hours deadline
http://prpnet.mine.nu:12000/all.html
But this port is set at 18 hrs
http://prpnet2.mine.nu:12000/all.html
They are both for the proth ports. Its just that when the second port is empty or my computer cant get WU from it it gets work from the first but as its an old computer its slow and will miss the dealine.
Can they both be 18 hrs?
Many Thanks
Ian
____________
Welcome to Holland
 |
|
|
|
|
|
I shall change that.
Lennart |
|
|
|
|
I shall change that.
Lennart
Thank you Lennart
____________
Welcome to Holland
|
|
|
|
|
|
Pls check PPSElow and SGS ports. |
|
|
|
|
Pls check PPSElow and SGS ports.
Checked. Should work now.
Lennart |
|
|
|
|
|
Around 1/3 of all ports are down. |
|
|
Sysadm@NbgVolunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 5 Feb 08
Posts: 1233
ID: 18646
Credit: 917,556,934
RAC: 380,242
             
|
Around 1/3 of all ports are down.
seems to be a problem with prpnet.primegrid.com-ports;
prpnet.mine.nu is online on all ports
____________
Sysadm@Nbg
my current lucky number: 113856050^65536 + 1
PSA-PRPNet-Stats-URL: http://u-g-f.de/PRPNet/
 |
|
|
|
|
|
Restarted.
Lennart |
|
|
|
|
|
Wall-Sun-Sun is empty.
____________
|
|
|
|
|
Wall-Sun-Sun is empty.
PPSElow2 too, Lennart, you can remove this port now ;) |
|
|
|
|
|
Ports 12000, 12001, 12005, & 12006 are't responding, looks like something needs another restart.
____________
147*2^1392930+1 was my first prime number found, others have followed :) |
|
|
|
|
|
Always the same ports are down, it looks like more unstable last time. |
|
|
Scott BrownVolunteer moderator
Project administrator
Volunteer tester
Project scientist
Send message
Joined: 17 Oct 05
Posts: 2416
ID: 1178
Credit: 19,978,285,056
RAC: 19,305,139
|
|
Lennart is aware of the issue and is looking into a solution.
____________
141941*2^4299438-1 is prime!
|
|
|
|
|
|
I got a strange error but I can't get it anymore :/
I have started all again. I hope it will work.
Lennart
|
|
|
|
|
I got a strange error but I can't get it anymore :/
I have started all again. I hope it will work.
Lennart
Nope, all prpnet.primegrid.com down again since days. Its better to port the rest of SGS to BOINC and close it, also PPSElow2 could be closed. |
|
|
|
|
|
ports of prpnet.primegrid.com down |
|
|
|
|
|
down? i got the message that the domain doesn't exist :-(
____________
DeleteNull |
|
|
|
|
|
Test now. It should work.
Lennart |
|
|
|
|
|
It works now, thank you. I could download and install the software. The only problem was to install vnc on the remote machine...
Test now. It should work.
Lennart
____________
DeleteNull |
|
|
|
|
|
I think SGS and GFN524288 can be removed from prpnet. Both are running better with BOINC. Should also spend free ressources and money. |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 14036
ID: 53948
Credit: 475,998,733
RAC: 222,579
|
I think SGS and GFN524288 can be removed from prpnet. Both are running better with BOINC. Should also spend free ressources and money.
Right now, GFN524288 isn't seeing much use (it may be completely dry), but I expect that to change in a few months. It's being crunched at a fantastic rate on BOINC, and it won't be long until we exhaust the CUDA-capable candidates. At that point, I'd like to move it back to PRPNET for further crunching using Genefer80 on CPUs. Over the last few months we did a lot of sieving on N=524288 for that purpose.
When that happens, the short BOINC tasks will switch to the next higher range of n=20 (N=1048576) which, as far as I know, has never been crunched by anyone before.
____________
My lucky number is 75898524288+1 |
|
|
