Join PrimeGrid
Returning Participants
Community
Leader Boards
Results
Other
drummers-lowrise
|
Message boards :
Project Staging Area :
Project Staging Area (Archive)
Author |
Message |
John Honorary cruncher
 Send message
Joined: 21 Feb 06 Posts: 2875 ID: 2449 Credit: 2,681,934 RAC: 0
                 
|
Welcome to the Project Staging Area otherwise known as Not Ready for Prime Time! :)
We recently announced some Upcoming Projects. While most of the prep work has been done internally, we wish to open this process to anyone who is interested in helping out.
Most of our resources are devoted to sieving as well they should be. However, with the +1 Prime Search (PPS), we find ourselves needing to do a lot of primality testing. BOINC LLR has a WU processing threshold of about 6-7 seconds. Therefore, the WU's need to take longer than that for BOINC LLR to function properly. We have set the minimum threshold at 10 seconds. However, while technically feasible, it definitely is not optimal. The load on the database alone would be overwhelming.
Therefore, we have increased that threshold to a minimum of 20 seconds with a preference towards a minute. For the PPS project, this equates to completing all n<150K with a preference towards n<200K. A C2Q 2.4 was used for timing purposes to be on the safe side.
We have completed n<100K and are now looking for help to complete n<150K and possibly n<200K. To participate, download LLRNet, a "semi-manual" program that acts a lot like BOINC. Once set up, you'll receive WU's and send results automatically.
Our primary focus is on PPS. However, other projects are available as well. If you wish to participate in these, please feel free to do so. For more information about these projects, please see the Upcoming Projects thread.
NOTE: Early in a projects life cycle, the work is mostly double checking previous results. Any new primes found (excluding the Sophie Germain Prime Search) will not make it into the Top 5000 primes database. Additionally, no stats or BOINC credit are planned at this time.
If you wish to participate, please see the next post. As these projects move into PrimeGrid BOINC production, we will continue to use this process to introduce future projects.
Thank you for your help.
____________
| |
|
John Honorary cruncher
 Send message
Joined: 21 Feb 06 Posts: 2875 ID: 2449 Credit: 2,681,934 RAC: 0
                 
|
Welcome to PrimeGrid LLRNet
NOTE: We will be transistioning to PRPNet from LLRNet. Please see this thread for more information: Beta test PRPNet
We'll be using LLRNet to help automate the primality tests. It is very simple to set up. Just follow these instructions and you'll be up and running in less than 5 minutes. :)
LLRNet is available for the following platforms:
LLRNet - Original Version 0.9b7
Windows
Linux
PrimeGrid LLRNet with Batch Jobs - TESTING - version PG-2.4
Windows
Linux
Version number can be found in the first line when LLRNet starts OR if you open lrnet.lua, it's on line 13 in the batch version and line 11 in the original version.
If you already have LLRNet, then all you need to update are the following three files: update_file.zip
llrnet.lua
client.lua
llr-clientconfig.txt
NOTE: You have to use these clients to do work on PrimeGrid LLRNet server!!! Because we test for other bases, the base 2 only version with GUI will not work.
After downloading and unpacking, the program is ready to run by simply double clicking llrnet.exe. However, if you wish to personalize your work, edit the llr-clientconfig.txt file. Here you can change the username from the default "pg-user" to your PrimeGrid username. It would be a good idea to read through the comments in the file and make changes to the sections you want. All is pretty much self-explanatory, but if you need any help, please do not hesitate to ask.
When started, a 4 vertical line task icon will show up in your task tray. Here you can "Open" a console to see the work you are doing by right clicking the icon. You would "Detach" (close) the console the same way. Also, please use this method when "Exiting" (shutting down) LLRNet.
Congratulations! You are now a member of the Not Ready for Prime Time Team.
Thank you!
Current Projects
server = "pgllr.mine.nu"
port 443: redirects to port 500
port 500: Sophie Germain Prime Search (SGS) (k*2^666666-1)
port 5000: Generalized Woodall; b=13 (n*b^n-1)
port 6333: Generalized Cullen; b=13 (n*b^n+1)
DEADLINE for all ports: 4 * 24 * 3600 (4 days)
Additional Information
- Seeing the Primes you have found - to see any primes that you have found, simply look in the prime.txt or primetime.txt files. They are automatically generated when you find a prime. The prime.txt file is used to easily copy the k n pair over to other programs for further testing. The primetime.txt file offers more detailed information about the prime found.
- Seeing the WU's you have completed - to see any WU's that you have completed, simply look in the lresults.txt file. It is automatically generated when you complete a WU. To see WU's that your computer has downloaded and will be working on, look in the workfile.txt file.
- Work Unit Deadline - currently all projects have a WU deadline of 48 hours. If not returned within that time, the WU is cancelled and sent out again. If you are not running PG_LLRNet full time, please keep a low cache level (around 5 is good). If you have WU's on your machine that are past the deadline, please see "Canceling Work Units."
- Cancelling Work Units - if you wish to cancel your WU's, please run endjob.bat and follow the on-screen instructions. This will cancel all WU's you have as well as delete the last checkpoint file. If you are using the batch version, just edit the llr-clientconfig.txt file and set canceljob to 1 or 2.
- Emptying your cache (without cancelling) - to empty your cache of all your WU's, edit the llr-clientconfig.txt and set your cache to zero (WUCacheSize = 0). Stop LLRNet and restart for changes to take effect. LLRNet will now complete each WU in your cache until it is empty. Once cache is empty, LLRNet will continue to download 1 WU at a time. To complete this last WU and not request any more, please see "Completing the last WU". If you are using the batch version, just set refill to 0.
- Completing the last WU - after setting your cache to zero and emptying all WU's, edit llr-clientconfig.txt and uncomment "-- once = 1" by removing "--". This will tell LLRNet to complete its current WU and close. Stop LLRNet and restart for changes to take effect. LLRNet will now complete the last WU and close automatically.
- Windows Vista Users - it seems that LLRNet may behave a bit odd running within Vista. Sometimes the LLRNet icon in the system tray will take the shape of the last icon Vista had open. Also, the options may appear blank when you right click it. They are as follows from top to bottom: 1. Exit LLRNet; 2. Open/Close console; 3. Add LLRNet as a service; 4. Hide icon in system tray. As for processing, LLRNet runs perfectly well on Vista. USER SUGGESTION: My latest trick for always making the icon to show is to start task manager (ctrl+shift+esc) first. While that is running, start, llrnet and the right icon will show.
- Batching Work Units - PrimeGrid's verion of LLRNet actually batches WU's and only connects to the server to request more when there are 2 WU's left. If you have 2 or less WU's set for your WUCacheSize, then it will run as the original version, connecting to the server after each completed WU.
NOTE for those participating in multiple projects or changing from one port to another
Changing port numbers (server name, username) while you still have work in your cache will cause the server not to recognize your results.
The simplest solution is to have a folder for each project/port. Otherwise, in order to change to other ports, you must first empty your cache. A simple example for 3 folders: PGLLRNet300, PGLLRNet500, and PGLLRNet5000. Then edit the llr-clientconfig.txt in each folder and change the port number to match your folder.
If you want, you can edit the "servicename" as well to match your folder. That way, if you have a multiple core machine and are running multiple projects at the same time, you can determine which tray icon goes with which project.
If you are running multiple cores on the same project, it is recommended to change folder and servicename according. Example: PGLLRNet300_1, PGLLRNet300_2, PGLLRNet300_3, and so on.
____________
| |
|
John Honorary cruncher
 Send message
Joined: 21 Feb 06 Posts: 2875 ID: 2449 Credit: 2,681,934 RAC: 0
                 
|
A new lua is available for those wishing to batch their jobs. This is especially useful when participating in the PPS project. All the other projects have a longer "testing" time.
This lua has the following features:
- WUCacheSize - same as the original except LLRNet will test the entire WUCacheSize before connecting to the server again. So if it's set at 20, LLRNet will test 20 WU's and then connect the server to download 20 more.
NOTE: WUCacheSize is hard coded to a max of 100 WU's. If you wish to increase this, edit the llrnet.lua line 56. Please notify Lennart when you make the change.
- refill - use this option to set a refill level. For example, if you have WUCacheSize = 50 and refill = 5, LLRNet will test 45 WU's and when the queue reaches 5 left, it will download another 45 WU's to return to the WUCacheSize = 50. Another nice option for refill is to set it to 0 if you want to run your cache down to 0 and exit.
- WUCacheSize = 0, refill = 0 - WUCacheSize and refill are set to 0. Therefore, no work will be downloaded. In order to receive work, please enter a value of at least 1.
- WUCacheSize = 1, refill = 0 - This is the same as the old version WUCacheSize = 0, once = 1. LLRNet will download 1 WU and close.
- WUCacheSize = 1, refill = 1 - This is the same as the old version WUCacheSize = 0. LLRNet will download 1 WU, test it, return it, and download another.
- WUCacheSize = 1, refill = 2 - This will error out...Your refill number is larger than your WUCacheSize. Please edit the llr-clientconfig.txt file to resolve this issue.
- canceljob - this option allows you to cancel your work. You don't have to run the endjob.bat batch job anymore. canceljob = 1 will cancel the current WU. canceljob = 2 will cancel the current job and all work in your cache.
- Console opens on start - the console will automatically open when started. If you wish for this to remain like the original version, just comment out lines 40-42 in the llrnet.lua file by preceding the lines with --.
- Screen output - this version outputs to the screen information of what is happening with the batching of WU's.
- Uploading WU's - in general, if there is completed work in the tosend.txt file when LLRNet starts, it will be uploaded first.
Please see second post in this thread for the latest version.
If you have any questions, please PM me or post in the LLRNet discussion thread. All feedback is welcomed.
John
____________
| |
|
John Honorary cruncher
 Send message
Joined: 21 Feb 06 Posts: 2875 ID: 2449 Credit: 2,681,934 RAC: 0
                 
|
PG-2.4 is availabe for those using the batching version. This resolves a small bug related to the WUCacheSize = 1, refill = 0 setting.
Also, for WU's that take longer than one minute, it supresses the extra output created by the batching information.
Please see second post for the files. For this update, all you need is the llrnet.lua. It's in the update_file.zip file. Just replace your current one with this one.
____________
| |
|
|
What's up with the maximum for wccachesize? It doesn't seem to be documented, but if I set it to 200 with a refill of 100, if says the refill is higher than the cache...
____________
| |
|
|
What's up with the maximum for wccachesize? It doesn't seem to be documented, but if I set it to 200 with a refill of 100, if says the refill is higher than the cache...
Yeah I've been curious about that also
____________
John M. Johnson "Novex" | |
|
|
What's up with the maximum for wccachesize? It doesn't seem to be documented, but if I set it to 200 with a refill of 100, if says the refill is higher than the cache...
There is a max 100 in llrnet.lua "WUCacheMaxSize = 100"
If you change that to 200 it should work
/Lennart
____________
| |
|
|
Question: How can I make LLRNet to use all four cores of my Q6600 machine(linux)?
Thanks,
Hugo | |
|
|
Question: How can I make LLRNet to use all four cores of my Q6600 machine(linux)?
Thanks,
Hugo
I have a folder named llrnet then a made llrnetet/l1
llrnet/l2
llrnet/l3
llrnet/l4
I install all on llrnet/l1
and copy all from llrnet/l1 to l2 l3 & l4
Then i open one terminal on each folder (l1 l2 l3 l4)
./llrnet in all and you have 4 running.
/Lennart | |
|
warddr Volunteer tester
 Send message
Joined: 7 Feb 08 Posts: 254 ID: 18735 Credit: 24,054,820 RAC: 0
           
|
what should be the username?
E-mailaddres or nickname (Ward De Ridder in my case) ore something else?
Ward | |
|
|
what should be the username?
E-mailaddres or nickname (Ward De Ridder in my case) ore something else?
Ward
Nick name works good..
/Lennart | |
|
warddr Volunteer tester
 Send message
Joined: 7 Feb 08 Posts: 254 ID: 18735 Credit: 24,054,820 RAC: 0
           
|
Does anyone know how to run llrnet on ubuntu?
./llrnet doesn't work,
make llrnet gives a lot of errors, anyone any suggestion, I installed build-essentials.
To make some things clear; the translation of
Bestand of map bestaat niet
is
File or folder doesn't exist
Errors:
warddr@warddr-laptop:~/Bureaublad/PG_LLRNetlinux$ make llrnet
gcc -DTHREADING "-DDEFAULT_LUA_FILE="llrnet.lua"" -g -c -Ilua/include -I../llrsource -I../llrsource/linuxllr -malign-double -o llrnet.o llrnet.cxx
llrnet.cxx:420: let op: deprecated conversion from string constant to ‘char*’
gcc -DTHREADING "-DDEFAULT_LUA_FILE="llrnet.lua"" -g -c -Ilua/include -I../llrsource -I../llrsource/linuxllr -malign-double -o net.o net.cxx
gcc -DTHREADING "-DDEFAULT_LUA_FILE="llrnet.lua"" -g -c -Ilua/include -I../llrsource -I../llrsource/linuxllr -malign-double -o threads.o threads.cxx
gcc -DTHREADING "-DDEFAULT_LUA_FILE="llrnet.lua"" -g -c -Ilua/include -I../llrsource -I../llrsource/linuxllr -malign-double -o llr2.o llr2.c
llr2.c:21:29: fout: linuxllr/lprime.h: Bestand of map bestaat niet
llr2.c:101:20: fout: cpuid.c: Bestand of map bestaat niet
llr2.c:102:19: fout: gwnum.c: Bestand of map bestaat niet
llr2.c:104:20: fout: Jacobi.c: Bestand of map bestaat niet
llr2.c:105:20: fout: giants.c: Bestand of map bestaat niet
llr2.c:106:22: fout: Igjacobi.c: Bestand of map bestaat niet
llr2.c:107:20: fout: Riesel.c: Bestand of map bestaat niet
llr2.c:108:17: fout: Llr.c: Bestand of map bestaat niet
llr2.c: In functie ‘sigterm_handler’:
llr2.c:114: interne fout ‘TRUE’ undeclared (first use in this function)
llr2.c:114: interne fout (Each undeclared identifier is reported only once
llr2.c:114: interne fout for each function it appears in.)
llr2.c: In functie ‘escapeCheck’:
llr2.c:220: interne fout ‘TRUE’ undeclared (first use in this function)
llr2.c:222: interne fout ‘FALSE’ undeclared (first use in this function)
llr2.c: In functie ‘SetPriority’:
llr2.c:249: interne fout ‘PRIORITY’ undeclared (first use in this function)
llr2.c: In functie ‘linuxContinue’:
llr2.c:291: interne fout ‘INI_FILE’ undeclared (first use in this function)
llr2.c:298: interne fout ‘_O_RDONLY’ undeclared (first use in this function)
llr2.c: In functie ‘isHighResTimerAvailable’:
llr2.c:376: interne fout ‘TRUE’ undeclared (first use in this function)
llr2.c:381: interne fout ‘FALSE’ undeclared (first use in this function)
llr2.c: In functie ‘primeTest’:
llr2.c:416: interne fout ‘giant’ undeclared (first use in this function)
llr2.c:416: interne fout expected ‘;’ before ‘gk’
llr2.c:417: interne fout expected ‘;’ before ‘gkk’
llr2.c:487: interne fout ‘gk’ undeclared (first use in this function)
llr2.c:573: interne fout ‘gkk’ undeclared (first use in this function)
llr2.c: In functie ‘lua_disableSSE2’:
llr2.c:811: interne fout ‘CPU_FLAGS’ undeclared (first use in this function)
llr2.c:811: interne fout ‘CPU_SSE2’ undeclared (first use in this function)
make: *** [llr2.o] Fout 1
warddr@warddr-laptop:~/Bureaublad/PG_LLRNetlinux$
____________
| |
|
|
To make some things clear; the translation of
Bestand of map bestaat niet
is
File or folder doesn't exist
Next time, use:
env LANG=C make llrnet
____________
Try an alternative BOINC client!
| |
|
warddr Volunteer tester
 Send message
Joined: 7 Feb 08 Posts: 254 ID: 18735 Credit: 24,054,820 RAC: 0
           
|
Here is the error again:
warddr@warddr-laptop:~/Bureaublad/PG_LLRNetlinux$ sudo env LANG=C make llrnet
gcc -DTHREADING "-DDEFAULT_LUA_FILE="llrnet.lua"" -g -c -Ilua/include -I../llrsource -I../llrsource/linuxllr -malign-double -o llr2.o llr2.c
llr2.c:21:29: error: linuxllr/lprime.h: No such file or directory
llr2.c:101:20: error: cpuid.c: No such file or directory
llr2.c:102:19: error: gwnum.c: No such file or directory
llr2.c:104:20: error: Jacobi.c: No such file or directory
llr2.c:105:20: error: giants.c: No such file or directory
llr2.c:106:22: error: Igjacobi.c: No such file or directory
llr2.c:107:20: error: Riesel.c: No such file or directory
llr2.c:108:17: error: Llr.c: No such file or directory
llr2.c: In function 'sigterm_handler':
llr2.c:114: error: 'TRUE' undeclared (first use in this function)
llr2.c:114: error: (Each undeclared identifier is reported only once
llr2.c:114: error: for each function it appears in.)
llr2.c: In function 'escapeCheck':
llr2.c:220: error: 'TRUE' undeclared (first use in this function)
llr2.c:222: error: 'FALSE' undeclared (first use in this function)
llr2.c: In function 'SetPriority':
llr2.c:249: error: 'PRIORITY' undeclared (first use in this function)
llr2.c: In function 'linuxContinue':
llr2.c:291: error: 'INI_FILE' undeclared (first use in this function)
llr2.c:298: error: '_O_RDONLY' undeclared (first use in this function)
llr2.c: In function 'isHighResTimerAvailable':
llr2.c:376: error: 'TRUE' undeclared (first use in this function)
llr2.c:381: error: 'FALSE' undeclared (first use in this function)
llr2.c: In function 'primeTest':
llr2.c:416: error: 'giant' undeclared (first use in this function)
llr2.c:416: error: expected ';' before 'gk'
llr2.c:417: error: expected ';' before 'gkk'
llr2.c:487: error: 'gk' undeclared (first use in this function)
llr2.c:573: error: 'gkk' undeclared (first use in this function)
llr2.c: In function 'lua_disableSSE2':
llr2.c:811: error: 'CPU_FLAGS' undeclared (first use in this function)
llr2.c:811: error: 'CPU_SSE2' undeclared (first use in this function)
make: *** [llr2.o] Error 1
warddr@warddr-laptop:~/Bureaublad/PG_LLRNetlinux$
____________
| |
|
rogueVolunteer developer
 Send message
Joined: 8 Sep 07 Posts: 1255 ID: 12001 Credit: 18,565,548 RAC: 0
 
|
It appears the there is no available work for either Generalized Woodall or Generalized Cullen. What is the status of those projects? It also appears that this is based upon LLR 3.5, not LLR 3.7. Another question I have is why are these using LLRNet instead of BOINC? | |
|
|
why are these using LLRNet instead of BOINC?
Due to the heavy load on previous server LOL | |
|
warddr Volunteer tester
 Send message
Joined: 7 Feb 08 Posts: 254 ID: 18735 Credit: 24,054,820 RAC: 0
           
|
why are these using LLRNet instead of BOINC?
Due to the heavy load on previous server LOL
-To test projects who arn't stabile yet
-To be able to run it on computers where you haven't got a superuser account
-To go back to basic (this is a litle more work to setup,...)
-To double check
-To find other kind of primes (like Sophie Germain primes)
-...
____________
| |
|
Message boards :
Project Staging Area :
Project Staging Area (Archive) |