| Author |
Message |
|---|
John Honorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
                 
|
Since we are very close to the 5M mark, could you be more specific about the double-checking ranges...?
There are some interesting prime gaps below 3M. Those are the first places to be explored:
- 901173-1076441
- 1199930-1395688
- 1649228-1779748
- 1993429-2178109
- 2639439-3201175 (3M+ has already been search by PG)
These could be perfectly legit gaps but there's no way of telling without a proper DC effort.
Based on our initial research, we concluded that RS had reached n=3M in their first pass work (excluding a 3 outliers which they took higher) and had double checked all work up to n=2M (no proof though). We base this on the information that they were using their LLRnet to do first pass work and BOINC to do DC work.
When RS closed down, their LLRNet was processing at n~3M while BOINC was processing n~2M. Should PG's DC effort find a prime below n=2M, then that opens the doors for a full DC check of n<2M.
The entire 2M-3M range is suspect and ripe for a DC effort. Note that this range had a first pass in their LLRNet.
Passed 5m now?
Yes, 5M has indeed been passed. :)
____________
|
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
Passed 5m now?
Yes, 5M has indeed been passed. :)
DC work going out starting with the 901173<n<1076441 range.
____________
|
|
|
|
|
|
Iam on now with 4 cores and avx. I have seen some minutes tasks, doublechecks? |
|
|
HonzaVolunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1893
ID: 352
Credit: 3,142,312,174
RAC: 0
                            
|
Iam on now with 4 cores and avx. I have seen some minutes tasks, doublechecks?
I think you just grabbed fresh WUs for DC as John announced.
215443*2^901211-1 is not prime.
____________
My stats
Badge score: 1*1 + 5*1 + 8*3 + 9*11 + 10*1 + 11*1 + 12*3 = 186 |
|
|
|
|
|
In the rieselsieve project, I found 469949 · 2^1649228 - 1 in their double-check run. So I think we can assume the double check was performed at least that far as the project ran for quite some time after that. |
|
|
|
|
|
I say we just double check the entire range: 0M-3M. Then we know for sure, and it's all DCed and proven. Besides, how long could it take? We now only have to test 57 k values, and our computers are way faster. |
|
|
|
|
|
I'm getting some interesting variations in credit on these shorter WUs.
Varying between 10.70 for a 450 second run time to 136.33 on a 535 second WU.
____________
35 x 2^3587843+1 is prime! |
|
|
|
|
DC work going out starting with the 901173<n<1076441 range.
Hi John,
I've been crunching on this sub-project since November (except during challenges), primarily to elevate my silver badge to a gold one (should happen today), but I was thinking of leaving some cores here until we get through the DC.
Can you give an estimate as to how long you think that may take with the current level of support? I see that TRP LLR is up for the 5 day challenge in May, do you think we will be done by then?
Thanks,
Peter
____________
35 x 2^3587843+1 is prime! |
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
DC work going out starting with the 901173<n<1076441 range.
I've been crunching on this sub-project since November (except during challenges), primarily to elevate my silver badge to a gold one (should happen today), but I was thinking of leaving some cores here until we get through the DC.
Can you give an estimate as to how long you think that may take with the current level of support? I see that TRP LLR is up for the 5 day challenge in May, do you think we will be done by then?
It's hard to say right now. These tests are much smaller so we'll go through the lower ranges more quickly. The 2M-3M range should take some time though. We will probably reach that range in a few weeks. A better gauge will be to see where we are at the end of the month.
It also depends on the confidence level of the the previous results. If a prime is found below n=2M, that will lead to further DC'ing.
____________
|
|
|
|
|
|
Congrats TroubledBunny
I"ll also get my TRP gold today. This has been a focus project for me since I was in the top 15-20 on the old rs project before it went boinc. Hopefully we will blow thru this dc real fast. |
|
|
|
|
Congrats TroubledBunny, I"ll also get my TRP gold today.
Same to you SlangNRox!
I don't think I'll hang around on this sub-project though, the nice fat credits seem to have dried up on those DCs. Guess I'll for gold in another sub-project, Woodall next I think. |
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
DC work going out starting with the 901173<n<1076441 range.
Can you give an estimate as to how long you think that may take with the current level of support?
It also depends on the confidence level of the the previous results. If a prime is found below n=2M, that will lead to further DC'ing.
Confidence of previous results has dropped. :( We'll continue to run through the gaps while deciding what to do next for n<3M DC'ing.
162941*2^993718-1 is prime! (299145 digits)
56 k's now remain.
____________
|
|
|
 Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 237,712,514
RAC: 0
                         
|
DC work going out starting with the 901173<n<1076441 range.
Can you give an estimate as to how long you think that may take with the current level of support?
It also depends on the confidence level of the the previous results. If a prime is found below n=2M, that will lead to further DC'ing.
Confidence of previous results has dropped. :( We'll continue to run through the gaps while deciding what to do next for n<3M DC'ing.
162941*2^993718-1 is prime! (299145 digits)
56 k's now remain.
I think this is probably the first time anyone ever announced knocking a K off of either the Riesel or Sierpinski problems -- and included a sad-face emoticon in the same post!
____________
My lucky number is 75898524288+1 |
|
|
|
|
|
Please, oh, please, can we just double check every N<3M, so we're sure? This new prime is so small! I have no confidence in the previous project whatsoever. A thorough DC effort won't take long for what it's worth, and then we'll have proof. I don't want to test N>5M without a full DC of N<3M.
Keep in mind, for the remaining 56k values, there is a very large gap of 0M-3M. That's pretty big!
If a full DC effort is started, I promise to move all of my computers back to TRP. I have completed 2732 tasks for 1,305,570.52 credit. I will certainly help out and do my part.
In any case, it's pretty awesome that PrimeGrid found an 8th prime for TRP! |
|
|
RytisVolunteer moderator
Project administrator
Send message
Joined: 22 Jun 05
Posts: 2649
ID: 1
Credit: 26,363,112
RAC: 0
                 
|
|
We're going to run through the ranges that we have highest doubt in first. We'll see about doublechecking remaining numbers once we know more about the outcome of the ranges we're going through now.
____________
|
|
|
|
|
|
Wow, this is a surprise to me, both good and bad. I do know they ran double checks previously, at least at lower levels, so how this was missed... |
|
|
|
|
We're going to run through the ranges that we have highest doubt in first. We'll see about doublechecking remaining numbers once we know more about the outcome of the ranges we're going through now.
I'm ready to throw all my cores in as well if we're going to double check the lower range. Eager to help eliminate more k's.
Cheers. |
|
|
RytisVolunteer moderator
Project administrator
Send message
Joined: 22 Jun 05
Posts: 2649
ID: 1
Credit: 26,363,112
RAC: 0
|
|
All the work that is going out now (unless resends happen) is going out in the n range below 3M.
____________
|
|
|
|
|
We're going to run through the ranges that we have highest doubt in first. We'll see about doublechecking remaining numbers once we know more about the outcome of the ranges we're going through now.
Rytis, should we temporarily put TRP Sieve on hold since we don't know which k's will be eliminated next? Otherwise it might turn out to be unnecessary waste of CPU power.
|
|
|
|
|
We're going to run through the ranges that we have highest doubt in first. We'll see about doublechecking remaining numbers once we know more about the outcome of the ranges we're going through now.
Rytis, should we temporarily put TRP Sieve on hold since we don't know which k's will be eliminated next? Otherwise it might turn out to be unnecessary waste of CPU power.
That is not needed !
To remove one or 2 k from sievefile will have a very little impact on the time to sieve.
Lennart |
|
|
RytisVolunteer moderator
Project administrator
Send message
Joined: 22 Jun 05
Posts: 2649
ID: 1
Credit: 26,363,112
RAC: 0
|
|
We have decided to start full doublecheck, so the server will start feeding work from n=800000. Lower ranges will be checked in PRPnet.
____________
|
|
|
|
|
|
I switched all of my computers to TRP! |
|
|
|
|
|
The timing is terrible or should I say unfortunate as there is also a push on the pps side to get to 1.5m. |
|
|
|
|
The timing is terrible or should I say unfortunate as there is also a push on the pps side to get to 1.5m.
Not really, as there always some PPS LLR push going on, especially as being the focus project. Of course, you could hope that people move here are leave the primes for you over there. Not to mention, you do have an extra day this year for the Tour! |
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
Confidence of previous results has dropped. :( We'll continue to run through the gaps while deciding what to do next for n<3M DC'ing.
162941*2^993718-1 is prime! (299145 digits)
56 k's now remain.
Good news and bad news.
The good news: Confidence has been improved considerably but not completely. :) We still have interest in the prime-less gaps but the chance of finding a missing prime outside that is much lower now.
We've been informed that the prime found was in a k/n range that was searched by an individual prior to RS taking up the search. Therefore, RS was not directly in error here. At worse, it appears that maybe RS did not fully DC previous work before taking on that particular k.
It could very well be that we misunderstood what RS considered double checked. Maybe their claim was that they fully DC'd their LLRNet work. In light of this new information, it's a very reasonable presumption. Sadly, there's no way to know for sure what the true meaning was.
As mentioned above, we are still interested in the prime-less gaps. Also, we are interested in the low n work that was completed before RS took on the problem. I'm waiting for more information but looking at Wilfrid Keller's The Riesel Problem: Definition and Status I would presume that most of n<1M work was done by individuals. n>800k is being completed in BOINC which leaves n<800k for PRPNet. Please feel free to join in on the effort by adding this server/port to your master_prpclient file and updating:
server=TRPDC:100:5:prpnet.mine.nu:14000 Progress is available here.
Now the bad news: There appears that some select k's were taken above n=1M AND it is unknown if RS DC'd them since they were not a part of their LLRNet effort. Therefore, a full DC is still in order, but don't count on finding many primes for the Tour. :) Also, the entire 2M-3M range was not DC'd by RS and the largest gap is there.
The timing is terrible or should I say unfortunate as there is also a push on the pps side to get to 1.5m.
Yes, getting to n=1.5M would make for a more "lively" Tour. And yes, while there's an extra day in February, the Riesel Problem will be around for a lot longer...certainly not going anywhere in our lifetimes. It's your choice what you wish to participate in. ;) NOTE: There's no Challenge Series in March. ;)
____________
|
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
|
More details have been provided about "take-over" points where RS most likely started their search from individual searchers. They are listed below, sorted by n, for the remaining 56 k's:
k n
40597 524288 --> RieselSieve
250027 524288 --> RieselSieve
344759 524288 --> RieselSieve
2293 524400 --> RieselSieve
121889 525000 --> RieselSieve
402539 525000 --> RieselSieve
384539 525600 --> RieselSieve
192971 528000 --> RieselSieve
365159 530000 --> RieselSieve
93839 556770 --> RieselSieve
362609 575000 --> RieselSieve
252191 580000 --> RieselSieve
129007 600000 --> RieselSieve
315929 600000 --> RieselSieve
368411 607150 --> RieselSieve
386801 660500 --> RieselSieve
319511 672140 --> RieselSieve
9221 710000 --> RieselSieve
23669 720000 --> RieselSieve
38473 720000 --> RieselSieve
67117 721900 --> RieselSieve
226153 723410 --> RieselSieve
234343 726000 --> RieselSieve
474491 729600 --> RieselSieve
215443 752210 --> RieselSieve
107347 767220 --> RieselSieve
325123 778500 --> RieselSieve
146561 804000 --> RieselSieve
97139 823490 --> RieselSieve
470173 833230 --> RieselSieve
477583 892930 --> RieselSieve
364903 905000 --> RieselSieve
273809 908150 --> RieselSieve
363343 909000 --> RieselSieve
304207 909600 --> RieselSieve
206231 911970 --> RieselSieve
494743 944390 --> RieselSieve
409753 949600 --> RieselSieve
502573 950000 --> RieselSieve
206039 999150 --> RieselSieve
485557 999980 --> RieselSieve
245561 1048500 --> RieselSieve
342847 1048500 --> RieselSieve
74699 1050000 --> RieselSieve
161669 1050000 --> RieselSieve
81041 1051200 --> RieselSieve
336839 1076940 --> RieselSieve
143047 1080000 --> RieselSieve
46663 1089000 --> RieselSieve
327671 1126100 --> RieselSieve
371893 1223000 --> RieselSieve
31859 1522000 --> RieselSieve
324011 1650000 --> RieselSieve
398023 1893000 --> RieselSieve
397027 1903800 --> RieselSieve
444637 2300000 --> RieselSieve
As suspected, most were before n=1M. The PRPNet effort should wrap up many of those ranges in the next few days. Starting at 800k, BOINC is continuing the double check. This will take longer as the deadline for a task is 5 days. It may take a few weeks before the trailing edge reaches n=1.5M. That will leave 5 remaining k's up to n=2.3M.
If there were any more "missteps" by individual searchers, we'll probably discover them in the next few weeks. Although a full DC effort is under way, the next prime-less gap to monitor is 1199930-1395688. Leading edge of the current search is at n=1159822.
____________
|
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
|
Here's a graph of 80 TRP primes found since 1998. The recent prime at n=993718 is included, closing the previous prime-less gap.
____________
|
|
|
|
|
|
Can the new port be added to http://u-g-f.de/PRPNet/? |
|
|
Sysadm@NbgVolunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 5 Feb 08
Posts: 1188
ID: 18646
Credit: 490,016,651
RAC: 0
               
|
|
I will do my very best ...
____________
Sysadm@Nbg
my current lucky number: 3749*2^1555697+1
PSA-PRPNet-Stats-URL: http://u-g-f.de/PRPNet/
 |
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
Can the new port be added to http://u-g-f.de/PRPNet/?
I will do my very best ...
This will be such a short lived port, it might be best to wait until it's completed and move it straight to ancient ports.
____________
|
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
|
Progress Update
There are some interesting prime gaps below 3M. Those are the first places to be explored:
- 901173-1076441
- 1199930-1395688
- 1649228-1779748
- 1993429-2178109
- 2639439-3201175 (3M+ has already been search by PG)
These could be perfectly legit gaps but there's no way of telling without a proper DC effort.
Trailing edge is at n=801431
Leading edge is at n=1308277
We are well within the next "prime-less" gap of 1199930-1395688.
____________
|
|
|
|
|
|
After a double check complete, do we check n>5M ? |
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
After a double check complete, do we check n>5M ?
Yes, progress will resume at n>5M.
____________
|
|
|
|
|
|
A question. Is it planned to DC also 'solved' k?
I mean, the Riesel Problem is 'for each k<509203 find an n such that k*2^n-1 is prime,'. Supposing that for each k we are checking increasing n, this means that the found n is also the smallest for that k, but this no longer holds. If we extend the problem to 'for each k<509203, find the smallest n such that k*2^n-1 is prime' this will need a DC also for solved k. |
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
A question. Is it planned to DC also 'solved' k?
I mean, the Riesel Problem is 'for each k<509203 find an n such that k*2^n-1 is prime,'. Supposing that for each k we are checking increasing n, this means that the found n is also the smallest for that k, but this no longer holds. If we extend the problem to 'for each k<509203, find the smallest n such that k*2^n-1 is prime' this will need a DC also for solved k.
Funny you should mention this. It just so happens that Wilfrid presented the same idea a few days ago and we're already working it. :) 'for each k<509203, find the smallest n such that k*2^n-1 is prime'.
PG has performed similar "min prime" checks before in the 5oB and extended Sierpinski problem projects. We would like to do the same for TRP. Once the PRPNet TRPDC has reached n=800k, we hope to load it with the remaining work to confirm that each listed prime is the min prime for that k. I believe it will start at n=262144 but that's not finalized yet. All k's with primes n<262144 are already in process.
[edit]
Not only will this process confirm min prime but it will also confirm the frequencies in the second table of Wilfrid Keller's site: The Riesel Problem: Definition and Status
____________
|
|
|
|
|
|
At first, I thought double-checking eliminated k values was silly, since the whole point is to merely eliminate them. However, accurately charting the frequencies of the "min primes" would provide useful information. |
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
|
PRPNet DC Progress
All work up to n=800k has been handed out. Only pendings remain outstanding. Once those are cleared, the server will be loaded with the minimum prime effort - 'for each k<509203, find the smallest n such that k*2^n-1 is prime'.
The minimum prime data is still being prepared...time consuming work. :) We hope to have something load within a day.
[edit] The TRPDC effort for the remaining 56 prime-less k's is complete to n=800k. No further primes were discovered.
____________
|
|
|
|
|
|
Very nice work. As someone who pushes TRP a lot, I appreciate it. TRP is my favorite on PrimeGrid.
____________
15547296^32768 + 1 is prime!
63 · 2^1356980 + 1 is prime!
63 · 2^1356980 + 1 divides xGF(1356973,11,4)! |
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
|
Progress Update
There are some interesting prime gaps below 3M. Those are the first places to be explored:
- 901173-1076441
- 1199930-1395688
- 1649228-1779748
- 1993429-2178109
- 2639439-3201175 (3M+ has already been search by PG)
These could be perfectly legit gaps but there's no way of telling without a proper DC effort.
Trailing edge is at n=801432
Leading edge is at n=1454556
Leading edge is beyond the "prime-less" gap of 1199930-1395688. Next in line is 1649228-1779748.
____________
|
|
|
NeoVolunteer tester
Send message
Joined: 28 Oct 10
Posts: 710
ID: 71509
Credit: 91,178,992
RAC: 0
          
|
|
Is there a new discovery?
501209*2^10008-1 3019 digits.
Found by KD7LRJ
|
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
Is there a new discovery?
Please see this post.
n<800k for the 56 prime-less k's is complete. No new discoveries.
We have entered into the minimum prime verification phase. Please see this post for more information.
____________
|
|
|
|
|
|
When I check out the subproject status, I notice k=67,117 has a "min in progress" of 368,605.
http://www.primegrid.com/stats_trp_llr.php
Every other k is above 800,000. |
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
|
PRPNet DC Progress
The TRPDC effort for the remaining 56 prime-less k's is complete to n=800k. No further primes were discovered.
The minimum prime verification has begun. Not only will this process confirm min prime but it will also confirm the frequencies in the second table of Wilfrid Keller's site: The Riesel Problem: Definition and Status
The TRPDC server has been loaded with 379 k's remaining at n=10k. The server stats page should light up like a switchboard as primes will come fast and furious. Once a prime is found for a particular k, that k is no longer searched. We are only looking for minimum primes.
The search will continue until all primes have been verified. We'll load additional work as it becomes available.
Thank you for participating.
____________
|
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
When I check out the subproject status, I notice k=67,117 has a "min in progress" of 368,605.
http://www.primegrid.com/stats_trp_llr.php
Every other k is above 800,000.
Yes, there was a hiccup in WU generation a few days ago and k=67117 released low n work. There are only two tasks outstanding below n=800k for that k.
____________
|
|
|
HonzaVolunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1893
ID: 352
Credit: 3,142,312,174
RAC: 0
|
|
Before I got a chance to join, PRPNet ort is empty.
____________
My stats
Badge score: 1*1 + 5*1 + 8*3 + 9*11 + 10*1 + 11*1 + 12*3 = 186 |
|
|
|
|
|
is the double checking done now?
Or will the port be refilled during the day/week?
____________
Member of the Dutch Power Cows
My Stats |
|
|
Sysadm@NbgVolunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 5 Feb 08
Posts: 1188
ID: 18646
Credit: 490,016,651
RAC: 0
|
Can the new port be added to http://u-g-f.de/PRPNet/?
I will do my very best ...
This will be such a short lived port, it might be best to wait until it's completed and move it straight to ancient ports.
as Honza wrote
PRPNet port is empty.
because of several posts in this thread the conclusion comes to me the search via prpnet is done
so the stats was moved to ancient ports ...
____________
Sysadm@Nbg
my current lucky number: 3749*2^1555697+1
PSA-PRPNet-Stats-URL: http://u-g-f.de/PRPNet/
|
|
|
|
|
Can the new port be added to http://u-g-f.de/PRPNet/?
I will do my very best ...
This will be such a short lived port, it might be best to wait until it's completed and move it straight to ancient ports.
as Honza wrote
PRPNet port is empty.
because of several posts in this thread the conclusion comes to me the search via prpnet is done
so the stats was moved to ancient ports ...
It is working currently.
you pulled the plug too early
____________
 |
|
|
Sysadm@NbgVolunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 5 Feb 08
Posts: 1188
ID: 18646
Credit: 490,016,651
RAC: 0
|
Can the new port be added to http://u-g-f.de/PRPNet/?
I will do my very best ...
This will be such a short lived port, it might be best to wait until it's completed and move it straight to ancient ports.
as Honza wrote
PRPNet port is empty.
because of several posts in this thread the conclusion comes to me the search via prpnet is done
so the stats was moved to ancient ports ...
It is working currently.
you pulled the plug too early
shame on me - where is the unchange buttom ;)
____________
Sysadm@Nbg
my current lucky number: 3749*2^1555697+1
PSA-PRPNet-Stats-URL: http://u-g-f.de/PRPNet/
|
|
|
HonzaVolunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1893
ID: 352
Credit: 3,142,312,174
RAC: 0
|
|
Celebration was short - empty again.
____________
My stats
Badge score: 1*1 + 5*1 + 8*3 + 9*11 + 10*1 + 11*1 + 12*3 = 186 |
|
|
|
|
|
How much time is needed to load extra DC work? I am waiting! |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 237,712,514
RAC: 0
|
Can the new port be added to http://u-g-f.de/PRPNet/?
I will do my very best ...
This will be such a short lived port, it might be best to wait until it's completed and move it straight to ancient ports.
as Honza wrote
PRPNet port is empty.
because of several posts in this thread the conclusion comes to me the search via prpnet is done
so the stats was moved to ancient ports ...
You're WAY early on that.
The problem is the WUs are being processed at an astounding rate. They fill up the port -- an hour later it's empty. Fill it up again, and it empties out quickly again.
We're going through them so fast that in the course of less than an hour, N has grown so much that the WUs are twice as long now!
But that doesn't mean we're necessarily going any slower because K's are being elliminated fairly quickly too. Well, maybe not as quickly as N is going up.
Anyway, we appear to be at 100 K now, and we have to go to 800 K, so it's going to be a while...
____________
My lucky number is 75898524288+1 |
|
|
Sysadm@NbgVolunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 5 Feb 08
Posts: 1188
ID: 18646
Credit: 490,016,651
RAC: 0
|
|
stats for the port active again / scanning interval every half hour (minute 17 and 47)...
... sorry for the missunderstanding at my side - i am not a native speaker and not a maths prof :-)
____________
Sysadm@Nbg
my current lucky number: 3749*2^1555697+1
PSA-PRPNet-Stats-URL: http://u-g-f.de/PRPNet/
|
|
|
NeoVolunteer tester
Send message
Joined: 28 Oct 10
Posts: 710
ID: 71509
Credit: 91,178,992
RAC: 0
|
|
Lost approximately 800 completed LLR tests due to candidates not being found on the server. Server so busy and/or offline that ESP was my backup with a 0%. Well, those ESP w/u's took 4 hours each.
I stopped all 10 cores, returned what I had completed, abandoned the rest of them. If I work this port again, it will be with a w/u cache of 7, not 100. Adjustment should be made to the server for timing deadline and w/u pulldown.
Edit - wow. thirty seconds later it's fixed. Perhaps it was the 1200 aborted w/u's I just sent in.
Frazzled.
Neo |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 237,712,514
RAC: 0
|
Lost approximately 800 completed LLR tests due to candidates not being found on the server. Server so busy and/or offline that ESP was my backup with a 0%. Well, those ESP w/u's took 4 hours each.
I stopped all 10 cores, returned what I had completed, abandoned the rest of them. If I work this port again, it will be with a w/u cache of 7, not 100. Adjustment should be made to the server for timing deadline and w/u pulldown.
Edit - wow. thirty seconds later it's fixed. Perhaps it was the 1200 aborted w/u's I just sent in.
Frazzled.
Neo
I had the same problem, but not quite as bad...
I started with 250 WU per core -- which was fine when the deadline was 30 minutes and the WU took 5 seconds each.
But we went through the queue so quickly that the next batch had a much higher N and took longer... and the next batch took even longer.
Each time through, I lowered the number of WUs to stay within the deadline. I still lost a few of them. :(
I was about to drop down to 10 WUs when the deadline was increased to 3 hours, so I'm back up to 100 WUs again. The WUs are currently taking about 70 to 75 seconds. I may be able to do one more batch of 100 after these, but that may be cutting it close. I may drop to 75 for the next batch. Of course, it's possible the server will be out of work again before the current batch finishes.
____________
My lucky number is 75898524288+1 |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 237,712,514
RAC: 0
|
|
The TRPDC port is spitting out +1 numbers instead of -1 numbers, and they all have small factors (3).
Example:
[2012-02-06 23:57:57 EST] Server: TRPDC, Candidate: 345067*2^270529+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-06 23:57:57 EST] Server: TRPDC, Candidate: 39269*2^270536+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-06 23:57:57 EST] Server: TRPDC, Candidate: 450457*2^270809+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-06 23:57:58 EST] Server: TRPDC, Candidate: 398533*2^270811+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-06 23:57:58 EST] Server: TRPDC, Candidate: 502541*2^270882+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-06 23:57:58 EST] Server: TRPDC, Candidate: 659*2^270896+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-06 23:57:58 EST] Server: TRPDC, Candidate: 417643*2^270895+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-06 23:57:59 EST] Server: TRPDC, Candidate: 504613*2^270899+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-06 23:57:59 EST] Server: TRPDC, Candidate: 412717*2^270905+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-06 23:57:59 EST] Server: TRPDC, Candidate: 39269*2^270912+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-06 23:58:26 EST] Server: TRPDC, Candidate: 109897*2^271037+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-06 23:58:26 EST] Server: TRPDC, Candidate: 376993*2^271187+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-06 23:58:26 EST] Server: TRPDC, Candidate: 98939*2^271280+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-06 23:58:27 EST] Server: TRPDC, Candidate: 267763*2^271443+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-06 23:58:27 EST] Server: TRPDC, Candidate: 277153*2^271491+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-06 23:58:27 EST] Server: TRPDC, Candidate: 39269*2^271496+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-06 23:58:27 EST] Server: TRPDC, Candidate: 299617*2^271517+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-06 23:58:28 EST] Server: TRPDC, Candidate: 401617*2^271525+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-06 23:58:28 EST] Server: TRPDC, Candidate: 450457*2^271529+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-06 23:58:28 EST] Server: TRPDC, Candidate: 460139*2^271532+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:04:58 EST] Server: TRPDC, Candidate: 401617*2^283085+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:04:58 EST] Server: TRPDC, Candidate: 113983*2^283087+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:04:59 EST] Server: TRPDC, Candidate: 104917*2^283097+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:04:59 EST] Server: TRPDC, Candidate: 267763*2^283099+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:04:59 EST] Server: TRPDC, Candidate: 345067*2^283141+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:04:59 EST] Server: TRPDC, Candidate: 192089*2^283144+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:04:59 EST] Server: TRPDC, Candidate: 398533*2^283147+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:00 EST] Server: TRPDC, Candidate: 43541*2^283154+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:00 EST] Server: TRPDC, Candidate: 103259*2^283156+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:00 EST] Server: TRPDC, Candidate: 220033*2^283155+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:08 EST] Server: TRPDC, Candidate: 113983*2^283407+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:08 EST] Server: TRPDC, Candidate: 144643*2^283411+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:08 EST] Server: TRPDC, Candidate: 201193*2^283439+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:08 EST] Server: TRPDC, Candidate: 204223*2^283443+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:08 EST] Server: TRPDC, Candidate: 109897*2^283445+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:09 EST] Server: TRPDC, Candidate: 113983*2^283447+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:09 EST] Server: TRPDC, Candidate: 113983*2^283455+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:09 EST] Server: TRPDC, Candidate: 162941*2^283462+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:09 EST] Server: TRPDC, Candidate: 192089*2^283480+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:09 EST] Server: TRPDC, Candidate: 212893*2^283483+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:41 EST] Server: TRPDC, Candidate: 342673*2^284511+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:41 EST] Server: TRPDC, Candidate: 325627*2^284549+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:41 EST] Server: TRPDC, Candidate: 325627*2^284645+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:41 EST] Server: TRPDC, Candidate: 412717*2^284649+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:42 EST] Server: TRPDC, Candidate: 502541*2^284666+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:42 EST] Server: TRPDC, Candidate: 500621*2^284670+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:42 EST] Server: TRPDC, Candidate: 460139*2^284672+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:42 EST] Server: TRPDC, Candidate: 382691*2^284674+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:42 EST] Server: TRPDC, Candidate: 71009*2^284680+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:43 EST] Server: TRPDC, Candidate: 93997*2^284705+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:53 EST] Server: TRPDC, Candidate: 104917*2^284897+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:53 EST] Server: TRPDC, Candidate: 152713*2^284899+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:53 EST] Server: TRPDC, Candidate: 357491*2^285002+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:53 EST] Server: TRPDC, Candidate: 277153*2^285003+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:54 EST] Server: TRPDC, Candidate: 192089*2^285016+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:54 EST] Server: TRPDC, Candidate: 114487*2^285021+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:54 EST] Server: TRPDC, Candidate: 220033*2^285027+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:54 EST] Server: TRPDC, Candidate: 357659*2^285028+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:54 EST] Server: TRPDC, Candidate: 275293*2^285035+1 Program: llr.exe Residue: small_factor Time: 0 seconds
[2012-02-07 00:05:55 EST] Server: TRPDC, Candidate: 192089*2^285040+1 Program: llr.exe Residue: small_factor Time: 0 seconds
____________
My lucky number is 75898524288+1 |
|
|
|
|
|
How can so many PRPnet points be lost in a single day? Seems that something is amiss in the stats generation. I captured this about 9:40Z Feb 7th, from http://u-g-f.de/PRPNet/alluser_stats.php?proj=TRPDC
--Gary
user total achieved today progress yesterday yesterday progress progress since a week ago progress since a month ago
ardo 1,097,823 0 1,097,823 0 1,097,823 1,097,823
KD7LRJ 907,626 514,245 393,381 170,472 907,626 907,626
rebirther 804,694 0 804,694 0 804,694 804,694
Gary_Craig 627,998 0 627,998 0 627,998 627,998
Crun-chi 563,062 0 563,062 0 563,062 563,062
Tarmo_Ilves 426,383 -317,791 744,174 118,769 426,383 426,383
meilijo 421,329 -700,240 1,121,569 76,346 421,329 421,329
Caravaggio 391,291 -374,125 765,416 32,279 391,291 391,291
Zydor 356,791 303,255 53,536 53,536 356,791 356,791
sm5ymt 281,441 -13,328,047 13,609,488 0 281,441 281,441
EDIT:: I just ran a few units on this port. As of the 9:47Z (10:47:02 CET) update, my total is now about 3,000; I just lost 624,000 credits. :-(
____________
"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) |
|
|
|
|
How can so many PRPnet points be lost in a single day? Seems that something is amiss in the stats generation. I captured this about 9:40Z Feb 7th, from http://u-g-f.de/PRPNet/alluser_stats.php?proj=TRPDC
--Gary
user total achieved today progress yesterday yesterday progress progress since a week ago progress since a month ago
ardo 1,097,823 0 1,097,823 0 1,097,823 1,097,823
KD7LRJ 907,626 514,245 393,381 170,472 907,626 907,626
rebirther 804,694 0 804,694 0 804,694 804,694
Gary_Craig 627,998 0 627,998 0 627,998 627,998
Crun-chi 563,062 0 563,062 0 563,062 563,062
Tarmo_Ilves 426,383 -317,791 744,174 118,769 426,383 426,383
meilijo 421,329 -700,240 1,121,569 76,346 421,329 421,329
Caravaggio 391,291 -374,125 765,416 32,279 391,291 391,291
Zydor 356,791 303,255 53,536 53,536 356,791 356,791
sm5ymt 281,441 -13,328,047 13,609,488 0 281,441 281,441
EDIT:: I just ran a few units on this port. As of the 9:47Z (10:47:02 CET) update, my total is now about 3,000; I just lost 624,000 credits. :-(
Something weird is happening:(
____________
|
|
|
Sysadm@NbgVolunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 5 Feb 08
Posts: 1188
ID: 18646
Credit: 490,016,651
RAC: 0
|
|
seems to me like a server reset
because I am collecting data daily, I can split the table daily - then only data between midnight (CET) and the reset is lost.
may an admin clarify if it is a temporary issue or not ...
____________
Sysadm@Nbg
my current lucky number: 3749*2^1555697+1
PSA-PRPNet-Stats-URL: http://u-g-f.de/PRPNet/
|
|
|
|
|
seems to me like a server reset
because I am collecting data daily, I can split the table daily - then only data between midnight (CET) and the reset is lost.
may an admin clarify if it is a temporary issue or not ...
It looks like the stats from all wu's crunched before the current work was loaded to the server has been deleted. Almost as if a new port had been set over the old one.
On my own, I did over 3500 tasks today. But total tasks done today shown here http://u-g-f.de/PRPNet/proj_stats?proj=TRPDC is still negative (as I write) almost 3k... which I think means than the number of tasks done today is less than the work erased... |
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
|
Sysadm@Nbg wrote: seems to me like a server reset
There was an accidental server reset. Using Sysadm@Nbg's history, we'll do our best to recapture the previous results. They won't be exact...but they will be pretty close.
Michael Goetz wrote: The TRPDC port is spitting out +1 numbers instead of -1 numbers, and they all have small factors (3).
Another accident. Fortunately, it was caught very early. As can be seen, the work that did go out was quickly processed, so not much computing time was lost...other than the downtime it took to correct it. ;)
____________
|
|
|
|
|
|
If I remember correctly similar thing happened with wwww too when we did beta testing
____________
|
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
If I remember correctly similar thing happened with wwww too when we did beta testing
Not exactly. That was a server crash and ALL data was lost. Here, only User and Team data was reset. All other data is preserved. We'll do our best to restore the User/Team data.
____________
|
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
|
From n=400k, here are the expected primes for the remaining 66 k's. Any primes found before these n's will be a new minimum n:
k n k n k n
398533 419107 261221 689422 234847 1535589
416413 424791 204223 696891 110413 1591999
46271 428210 220033 719731 469949 1649228
299617 428917 212893 730387 357659 1779748
277153 429819 246299 752600 417643 1800787
382691 431722 460139 779536 345067 1876573
201193 457615 659 800516 467917 1993429
401617 470149 93997 864401 196597 2178109
465869 497596 170591 866870 114487 2198389
144643 498079 309817 901173 450457 2307905
43541 507098 162941 993718 275293 2335007
401143 532927 150847 1076441 26773 2465343
458743 547791 412717 1084409 342673 2639439
98939 575144 504613 1136459 113983 3201175
89707 578313 500621 1138518 191249 3417696
357491 609338 350107 1144101 428639 3506452
222997 613153 152713 1154707 485767 3609357
103259 615076 71009 1185112 65531 3629342
279703 616235 502541 1199930 415267 3771929
126667 626497 192089 1395688 123547 3804809
109897 630221 149797 1414137 141941 4299438
215503 649891 325627 1472117 353159 4331116
p.s. Since these are known primes and we're just verifying minimum n, they will not count towards the Tour. However, if a new minimum n is established and the prime is large enough to enter into the Top 5000, that prime will count. :)
____________
|
|
|
|
|
|
Just check one of my test_result.log ..
Started with test of 1-2 seconds.. and gradually increasing times.. now at around 400 seconds/450K iterations per WU.. (also seen some of 500+ seconds)
Top5000 only needs a prime of 209399 digits :)
Let's go for it :) :) :)
____________
Member of the Dutch Power Cows
My Stats |
|
|
|
|
From n=400k, here are the expected primes for the remaining 66 k's. Any primes found before these n's will be a new minimum n:
This means that we have to check each of these k from n=400000 to n=min(3M ; n_known for that k) ? (from first post it seems that 3M+ is already doublechecked in PG) |
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
From n=400k, here are the expected primes for the remaining 66 k's. Any primes found before these n's will be a new minimum n:
This means that we have to check each of these k from n=400000 to n=min(3M ; n_known for that k) ? (from first post it seems that 3M+ is already doublechecked in PG)
PG started it's search at n=3M and used Adaptive Replication (AR). In short, this means that not all WU's were double checked. Although highly unlikely, it's still plausible that a lower prime could exist if PG missed it with a single replication WU.
We looking into removing the WU's where AR did do a double replication. This will be for n=3M to n=prime.
____________
|
|
|
|
|
|
Previously loaded "old" results have a standard CPU run time of 600 seconds which is an unrealistic time regarding new longer WUs...
Could you adjust those times that granted credits will reflect the work done more correctly?
One example: http://www.primegrid.com/workunit.php?wuid=248654438
"llr_trp_40597_1.7M_120462286"
Took me about an hour (3600sec) in comparison to the standard set 600 seconds. |
|
|
RytisVolunteer moderator
Project administrator
Send message
Joined: 22 Jun 05
Posts: 2649
ID: 1
Credit: 26,363,112
RAC: 0
|
|
That CPU time is not taken into account for granting credit anymore. It should grant based on the average recent runtimes.
____________
|
|
|
|
|
|
Sry, my mistake, didn't know that.
One other question: What does the "1.7M" in the WU name mean?
Is it the approximate size of n being tested?
Edit: Can answer my own question:
llr_trp_234343_1.7M_120462278
--> 234343*2^1782191-1 is not prime. |
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
|
TRPDC Progress
Leading edge of the BOINC DC effort for the 56 remaining prime-less k's is now at n=1813270. Range stats are available here.
For the PRPNet minimum prime effort, n=600k has been reached. There are now 51 k's remaining to confirm minimum primes. Server stats are available here.
k n k n k n
357491 609338 162941 993718 234847 1535589
222997 613153 150847 1076441 110413 1591999
103259 615076 412717 1084409 469949 1649228
279703 616235 504613 1136459 357659 1779748
126667 626497 500621 1138518 450457 2307905
109897 630221 350107 1144101 275293 2335007
215503 649891 152713 1154707 26773 2465343
261221 689422 71009 1185112 342673 2639439
204223 696891 502541 1199930 113983 3201175
220033 719731 192089 1395688 191249 3417696
212893 730387 149797 1414137 428639 3506452
246299 752600 325627 1472117 485767 3609357
460139 779536 417643 1800787 65531 3629342
659 800516 345067 1876573 415267 3771929
93997 864401 467917 1993429 123547 3804809
170591 866870 196597 2178109 141941 4299438
309817 901173 114487 2198389 353159 4331116
NOTE: Since these are known primes and we're just verifying minimum n, they will not count towards the Tour. However, if a new minimum n is established and the prime is large enough to enter into the Top 5000, that prime will count. :)
Sieving continues for 26 of the 33 remaining k's with expected primes > 1M. The 7 k's that PG found primes for were already sieved quite well and were retrieved from previous sieve files.
Additionally, tests that were already DC'd in BOINC for those 7 k's will be removed from further testing. This will probably be accomplished within the week. There's no hurry since PG started at 3M and we still have a ways to go to get there. :)
Thank you to everyone participating in this effort.
____________
|
|
|
|
|
|
The last TRP prime my machine found during the Double Check effort was 1st reported to the Top5000 list over 7 years ago. It entered the list as the 62nd largest prime known at the time. Now, it ranks as #3506.
At the time, it took 7.6 hours to prove. My machine took only 13.2 minutes.
Amazing, the progress made! :)
   ...Doug
____________
 |
|
|
|
|
|
I've been running a couple cores from a laptop on the PRPnet TRP DC port off and on for the last week or so. Since I use the laptop for "real work" as well, I need to turn off the PRPnet clients during parts of the day when I'm carrying the computer around and running it on battery. However, I've noticed that my reserved candidates on the server keep expiring--somewhat sooner than they should, in fact, based on the expiry deadlines reported on the server.
As an example, I had the clients running (with a batch size of 10 each) earlier today up to about 2:00 PM EST. I then turned off the clients until about 5:30 PM or so when I got back home; yet when I did so, the results were rejected by the server. I checked ]http://prpnet.mine.nu:14000/all.html and saw that my candidates were no longer listed in the pending tests table, even though the deadline is 8 hours and it had been much less than that (only about 3.5-4 hours) since the clients had last communicated with the server. The laptop's CPU is a Sandy Bridge i7 dual-core and I'm running llrAVX, so it's quite fast and definitely wasn't having any problems finishing the tests in time. When I last checked it was taking about an hour to get through a batch of 10 candidates.
Does anyone know why this is happening? Are older tests perhaps being expired manually from the backend prior to the 8 hour deadline?
Thanks,
Max :-) |
|
|
rogueVolunteer developer
Send message
Joined: 8 Sep 07
Posts: 1218
ID: 12001
Credit: 18,565,548
RAC: 0
|
|
You'd have to ask Lennart to look at the server log. It will indicate when the server expired the workunit. Can you verify when your client was given the workunits vs. when it tried to send them back to the server? |
|
|
|
|
You'd have to ask Lennart to look at the server log. It will indicate when the server expired the workunit. Can you verify when your client was given the workunits vs. when it tried to send them back to the server?
Sure--here's a relevant excerpt from one of the client logs:
[2012-02-15 13:51:17 EST] PG_TRPDC: Getting work from server 10.23.102.251 at port 14000
[2012-02-15 13:51:18 EST] PG_TRPDC: INFO: Server has a limit of 10 work units.
[2012-02-15 13:51:20 EST] PG_TRPDC: PRPNet server is version 5.0.2
[2012-02-15 13:53:49 EST] PG_TRPDC: Could not open file [lresults.txt] for reading. Assuming user stopped with ^C
/* Note: this is where I stopped the client manually. */
[2012-02-15 13:53:49 EST] Total Time: 9:16:55 Total Work Units: 57 Special Results Found: 0
[2012-02-15 13:53:49 EST] Client shutdown complete
[2012-02-15 17:46:14 EST] PRPNet Client application v5.0.6 started
[2012-02-15 17:46:14 EST] User name mdettweiler at email address is max@noprimeleftbehind.net
[2012-02-15 17:52:06 EST] PG_TRPDC: 450457*2^864849-1 is not prime. Residue 7ED4FA19C03CE3A4
[2012-02-15 18:00:31 EST] PG_TRPDC: 275293*2^864851-1 is not prime. Residue 40E87F881701E37B
[2012-02-15 18:04:30 EST] PG_TRPDC: Could not open file [lresults.txt] for reading. Assuming user stopped with ^C
[2012-02-15 18:04:30 EST] Total Time: 0:18:23 Total Work Units: 2 Special Results Found: 0
[2012-02-15 18:04:31 EST] PG_TRPDC: Returning work to server 10.23.102.251 at port 14000
[2012-02-15 18:04:32 EST] PG_TRPDC: INFO: Test for 450457*2^864849-1 was ignored. Candidate and/or test was not found
[2012-02-15 18:04:32 EST] PG_TRPDC: INFO: Test for 275293*2^864851-1 was ignored. Candidate and/or test was not found
[2012-02-15 18:04:33 EST] PG_TRPDC: INFO: 0 of 2 test results were accepted
[2012-02-15 18:04:33 EST] Client shutdown complete
Please ignore the strange-looking IP address in place of the real server location--I'm behind a proxy server and use an SSH tunnel on one of my other computers to get through to the server. (The 10.23.102.251:14000 shown in the log is tunneled to prpnet.mine.nu:14000.) :-) As far as I know, this shouldn't have any bearing on the issue at hand. |
|
|
rogueVolunteer developer
Send message
Joined: 8 Sep 07
Posts: 1218
ID: 12001
Credit: 18,565,548
RAC: 0
|
|
That does seem to be a problem. We'll have to wait to see what Lennart has in the server log. Maybe they were manually expired. |
|
|
RytisVolunteer moderator
Project administrator
Send message
Joined: 22 Jun 05
Posts: 2649
ID: 1
Credit: 26,363,112
RAC: 0
|
|
One more prime found in the doublecheck effort. Stay tuned...
____________
|
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
One more prime found in the doublecheck effort. Stay tuned...
Now two primes...same k.
____________
|
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
One more prime found in the doublecheck effort. Stay tuned...
Now two primes...same k.
If you're thinking, WOW, too good to be true, you'd be right. :(
More details to come.
____________
|
|
|
|
|
|
No new primes? Not even one? |
|
|
HonzaVolunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1893
ID: 352
Credit: 3,142,312,174
RAC: 0
|
|
I would guess just one.
Two primes for the same k would be too good.
I guess there were already such tests done.
____________
My stats
Badge score: 1*1 + 5*1 + 8*3 + 9*11 + 10*1 + 11*1 + 12*3 = 186 |
|
|
|
|
|
I don't see how two primes for the same k would be much better than one prime for that k. We only need one prime to knock off the k value. A second prime is meaningless wrt the Riesel Problem. |
|
|
|
|
|
If it is smaller it could provide better statistics for prime distribution |
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
No new primes? Not even one?
Not even one. Details may still be few days out since it's the weekend.
____________
|
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
|
TRPDC Progress
Leading edge of the BOINC DC effort for the 56 remaining prime-less k's is now at n=2221304. Range stats are available here.
There have been 3 false hits in the past few days. We have not yet pinpointed the problem nor has dialog with the developers been established. We have provided as many details as we could gather and are now waiting on a response. As soon as we know, we'll post in the forum.
For the PRPNet minimum prime effort, n=600k has been reached. There are now 33 k's remaining to confirm minimum primes. Server stats are available here.
k n k n
150847 1076441 467917 1993429
412717 1084409 196597 2178109
504613 1136459 114487 2198389
500621 1138518 450457 2307905
350107 1144101 275293 2335007
152713 1154707 26773 2465343
71009 1185112 342673 2639439
502541 1199930 113983 3201175
192089 1395688 191249 3417696
149797 1414137 428639 3506452
325627 1472117 485767 3609357
234847 1535589 65531 3629342
110413 1591999 415267 3771929
469949 1649228 123547 3804809
357659 1779748 141941 4299438
417643 1800787 353159 4331116
345067 1876573
NOTE: Since these are known primes and we're just verifying minimum n, they will not count towards the Tour. However, if a new minimum n is established and the prime is large enough to enter into the Top 5000, that prime will count. :)
Sieving continues for 26 of the 33 remaining k's with expected primes > 1M. The 7 k's that PG found primes for were already sieved quite well and were retrieved from previous sieve files.
Additionally, tests that were already DC'd in BOINC for those 7 k's will be removed from further testing.
Thank you to everyone participating in this effort.
____________
|
|
|
|
|
No new primes? Not even one?
Not even one. Details may still be few days out since it's the weekend.
Are there any news on the discoveries? |
|
|
|
|
|
Is it possible to increase the time limit on the TRPDC wu's? I see a lot of wu's getting recycled because they time out.
Perhaps 6h or 8h would be better given that n is now at 1.4M and most people request 5 wu's at a time.
|
|
|
|
|
|
Changed
Lennart |
|
|
|
|
|
John, a find of a Riesel prime has been on my account page for a week already but nothing is on the primelist page. Could please update us on new findings?
Cheers |
|
|
|
|
I'm getting some interesting variations in credit on these shorter WUs.
Varying between 10.70 for a 450 second run time to 136.33 on a 535 second WU.
<obligatory credit whinge>
I've noticed the credit for these double checks is significantly lower than on other subprojects </obligatory credit whinge>
|
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
No new primes? Not even one?
Not even one. Details may still be few days out since it's the weekend.
Are there any news on the discoveries?
There were 3 discoveries that turned out to be false positives. While not unheard of, it is pretty rare to have such a close grouping of them. There is active research ongoing with the "major developers" into the root of the problem. Hopefully they'll be able to pinpoint the issue and identify a "narrow band" of candidates affected.
When more news is available, this thread will be updated.
____________
|
|
|
|
|
|
It looks like I have found a prime for k=428639 at a lower n than what John had posted earlier.
[2012-03-24 17:45:26 EDT] Server: TRPDC, Candidate: 428639*2^2093836-1 Program: llr Residue: PRIME Time: 11542 seconds
Another false positive?
|
|
|
|
|
It looks like I have found a prime for k=428639 at a lower n than what John had posted earlier.
[2012-03-24 17:45:26 EDT] Server: TRPDC, Candidate: 428639*2^2093836-1 Program: llr Residue: PRIME Time: 11542 seconds
Another false positive?
Another one found a prime for the same number but other n value. There must be something wrong. |
|
|
|
|
|
428639*2^2093836-1, iteration : 2090000 / 2093836 [99.81%]. Time per iteration
428639*2^2093836-1 is not prime. LLR Res64: DAA0C30BC4A020FF Time : 1998.966 s
ec.
C:\llr388dev64>
Please EVERYONE Download the new PRPNET package or download LLR version 3.8.8 !!!!!
HERE http://jpenne.free.fr/Development
LLR 3.8.8 support AVX and work on all other system also.
Do not use any old version !!
We are working on updates to Boinc. SGS LLR Linux app is allready online.
Lennart
|
|
|
|
|
|
I have double check the other one:
428639*2^2094676-1 is not prime. LLR Res64: 80A17F83646B03B8 Time : 3490.549 sec.
You can reset the prime flag in table.
@Lennart: Pls inform via rss about the important update of llr/pfgw. This info is a little bit hidden. |
|
|
|
|
|
Hi,
I can confirm both:
428639*2^2093836-1
and
428639*2^2094676-1
trigger the known bug in LLR 3.8.7/gwnum.
Please upgrade to LLR 3.8.8 (in PRPNet package 5.0.7).
Cheers
- Iain
____________
Twitter: IainBethune
Proud member of team "Aggie The Pew". Go Aggie!
3073428256125*2^1290000-1 is Prime! |
|
|
|
|
|
http://prpnet.mine.nu:14000/server_stats.html
The stats page still shows two 2 primes found for that k value. |
|
|
|
|
|
It appears that the double check process is complete!
Back to virgin territory...
All that's left is to double-check the "min prime" k values. 33 remaining k values...
http://prpnet.mine.nu:14000/server_stats.html |
|
|
|
|
|
Maybe at n=5M time-deadline of Riesel WU should be rised a bit.
____________
  |
|
|
|
|
|
Also it seems that new WU are not double checked, like instead PSPllr and SoBllr.
I think we shall double check also TRPllr, since we are now in 'virgin territory'
____________
|
|
|
|
|
|
TRP llr uses adaptive replication (AR)...
"Trusted" hosts don't need to have their work double-checked.
Double-checking (presumable) correct results would be a waste of time and energy...
More info:
http://boinc.berkeley.edu/trac/wiki/AdaptiveReplication |
|
|
pschoeferVolunteer developer
Volunteer tester
Send message
Joined: 20 Sep 05
Posts: 662
ID: 845
Credit: 2,220,370,221
RAC: 0
                 
|
TRP llr uses adaptive replication (AR)...
"Trusted" hosts don't need to have their work double-checked.
Double-checking (presumable) correct results would be a waste of time and energy...
More info:
http://boinc.berkeley.edu/trac/wiki/AdaptiveReplication
I don't think that AR is used right now. I haven't seen any WU with quorum 2, even on a machine that recently crashed some WUs due to a dying HDD and had not returned a valid result since then.
____________
  |
|
|
|
|
|
Any news? It seems that DC has restarted and is now at 1.1M?
____________
|
|
|
|
|
|
Is the prime for k=415267 just below n=3M real? There was one found for n=3771929 last year in May which would then not be the smallest one. |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 237,712,514
RAC: 0
|
Is the prime for k=415267 just below n=3M real? There was one found for n=3771929 last year in May which would then not be the smallest one.
It's being checked right now.
____________
My lucky number is 75898524288+1 |
|
|
ardo
Send message
Joined: 12 Dec 10
Posts: 168
ID: 76659
Credit: 1,690,471,713
RAC: 0
       
|
|
And we're also running low on WUs (only 133 to go with 44 in progress)...
____________
Badge score: 2*5 + 8*7 + 3*8 + 3*9 + 1*10 + 1*11 + 1*13 = 151
 |
|
|
|
|
Is the prime for k=415267 just below n=3M real? There was one found for n=3771929 last year in May which would then not be the smallest one.
It's being checked right now.
It looks like another false positive. I ran an llr test and it turns out to be composite:
415267*2^2989385-1 is not prime. LLR Res64: F521BE01284087EB Time : 15212.228 sec.
I used prpclient 5.0.7 and llr version 3.8.8
|
|
|
|
|
Is the prime for k=415267 just below n=3M real? There was one found for n=3771929 last year in May which would then not be the smallest one.
It's being checked right now.
It looks like another false positive. I ran an llr test and it turns out to be composite:
415267*2^2989385-1 is not prime. LLR Res64: F521BE01284087EB Time : 15212.228 sec.
I used prpclient 5.0.7 and llr version 3.8.8
Could be in use of an older bugged llr. |
|
|
ardo
Send message
Joined: 12 Dec 10
Posts: 168
ID: 76659
Credit: 1,690,471,713
RAC: 0
|
Is the prime for k=415267 just below n=3M real? There was one found for n=3771929 last year in May which would then not be the smallest one.
It's being checked right now.
It looks like another false positive. I ran an llr test and it turns out to be composite:
415267*2^2989385-1 is not prime. LLR Res64: F521BE01284087EB Time : 15212.228 sec.
I used prpclient 5.0.7 and llr version 3.8.8
Could be in use of an older bugged llr.
This host is running a freshly installed prpclient 5.0.8 and llr 3.8.9 on Win7 64-bit.
____________
Badge score: 2*5 + 8*7 + 3*8 + 3*9 + 1*10 + 1*11 + 1*13 = 151
|
|
|
ardo
Send message
Joined: 12 Dec 10
Posts: 168
ID: 76659
Credit: 1,690,471,713
RAC: 0
|
This host is running a freshly installed prpclient 5.0.8 and llr 3.8.9 on Win7 64-bit.
I just noticed it is running the 32-bit version iso the 64-bit version of llr...
____________
Badge score: 2*5 + 8*7 + 3*8 + 3*9 + 1*10 + 1*11 + 1*13 = 151
|
|
|
ardo
Send message
Joined: 12 Dec 10
Posts: 168
ID: 76659
Credit: 1,690,471,713
RAC: 0
|
And we're also running low on WUs (only 133 to go with 44 in progress)...
Shouldn't the k's for which there are currently no tasks available but for which no prime has been found yet also be refilled...???
____________
Badge score: 2*5 + 8*7 + 3*8 + 3*9 + 1*10 + 1*11 + 1*13 = 151
|
|
|
|
|
|
I will load the other k's also but some time later.
Lennart |
|
|
|
|
I will load the other k's also but some time later.
Lennart
will that include candidates for k=415267?
I ran a manual llr test on two machines for 415267*2^2989385-1 and it came back composite in both cases with same residual.
machine A: WinXP - prpclient 5.0.7 - llr 3.8.8
machine B: OSX - prpclient 5.0.8 - llr 3.8.9 |
|
|
|
|
|
Yes that k will also be loaded.
Lennart |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 237,712,514
RAC: 0
|
|
Ardo,
Is that false positive result reproducible with either the 32 bit or 64 bit version of LLR?
____________
My lucky number is 75898524288+1 |
|
|
ardo
Send message
Joined: 12 Dec 10
Posts: 168
ID: 76659
Credit: 1,690,471,713
RAC: 0
|
Ardo,
Is that false positive result reproducible with either the 32 bit or 64 bit version of LLR?
Let me try that later today...
____________
Badge score: 2*5 + 8*7 + 3*8 + 3*9 + 1*10 + 1*11 + 1*13 = 151
|
|
|
ardo
Send message
Joined: 12 Dec 10
Posts: 168
ID: 76659
Credit: 1,690,471,713
RAC: 0
|
Ardo,
Is that false positive result reproducible with either the 32 bit or 64 bit version of LLR?
I reran this with the 32-bit version of LLR and now it has the same residue as below.
Must have been a fluke...
____________
Badge score: 2*5 + 8*7 + 3*8 + 3*9 + 1*10 + 1*11 + 1*13 = 151
|
|
|
|
|
|
So what's the deal with the RP min-prime double check? It seems kind of messy at this point... |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 237,712,514
RAC: 0
|
So what's the deal with the RP min-prime double check? It seems kind of messy at this point...
Nah, it was just a single non-reproducible error on one computer. It's no different than any time you get "Inconclusive" on the BOINC side. Errors happen; That's why things are double checked. Don't forget, that port is entirely a double check of previous results. Everything should come up composite. If it doesn't, we check it on the servers to see which one was correct.
The big problem earlier this year was different; ALL the computers were falsely saying some numbers were prime because of a programming error in LLR. That's not the situation here.
)
____________
My lucky number is 75898524288+1 |
|
|
|
|
|
Well, I see that there are three remaining k values which need to be double-checked, but the numbers currently being tested are for k values with min primes above n=3M. Aren't all of the n>3M already double-checked?
Also, will the k=415267 be corrected, since 415267*2^2989385-1 is composite? |
|
|
|
|
|
When will this double check be finished? The page says that there are 8 k values remaining... |
|
|
ardo
Send message
Joined: 12 Dec 10
Posts: 168
ID: 76659
Credit: 1,690,471,713
RAC: 0
|
|
One more down (342673*2^2639439-1) and only five more to go.
____________
Badge score: 2*5 + 8*7 + 3*8 + 3*9 + 1*10 + 1*11 + 1*13 = 151
|
|
|
|
|
only five more to go
Good!
Originally, I thought this min-prime double check would have been finished back in June! hehe |
|
|
ardo
Send message
Joined: 12 Dec 10
Posts: 168
ID: 76659
Credit: 1,690,471,713
RAC: 0
|
|
And one more down (415267*2^3771929-1) and now only two more to go.
____________
Badge score: 2*5 + 8*7 + 3*8 + 3*9 + 1*10 + 1*11 + 1*13 = 151
|
|
|
|
|
|
Will "my" prime (252191*2^5497878-1) -- which is btw. still missing on the PrimeGrid front page in the list of significant primes -- also be included in the double check effort? |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 237,712,514
RAC: 0
|
Will "my" prime (252191*2^5497878-1) -- which is btw. still missing on the PrimeGrid front page in the list of significant primes -- also be included in the double check effort?
As I'm sure you know, it's not currently in the DC effort. I don't know if there's any plans to do so in the future.
Yes, it should be on the home page (and on the Wiki, if it's not already there). I'll make sure that happens.
EDIT: I added the prime to the home page and it was already on the wiki.
____________
My lucky number is 75898524288+1 |
|
|
|
|
|
Which k values were not double-checked, and why were they not double-checked? Just k = 252,191? |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 237,712,514
RAC: 0
|
Which k values were not double-checked, and why were they not double-checked? Just k = 252,191?
I don't think anyone is saying specific k's weren't doublechecked.
TRPDC is DCing work done prior to PrimeGrid taking over the project.
Every LLR done here on BOINC's main TRP project is doublechecked. Except when it isn't:
We have Adaptive Replication turned on, which means that normally every number needs to be checked twice, except that "reliable" hosts don't need to be doublechecked. Therefore, *some* composite results in our main TRP BOINC project are only single checked, but it's not done for specific k's.
Primes are always doublechecked.
____________
My lucky number is 75898524288+1 |
|
|
ardo
Send message
Joined: 12 Dec 10
Posts: 168
ID: 76659
Credit: 1,690,471,713
RAC: 0
|
|
And the last two (141941*2^4299438-1 and 353159*2^4331116-1) have fallen also.
____________
Badge score: 2*5 + 8*7 + 3*8 + 3*9 + 1*10 + 1*11 + 1*13 = 151
|
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 237,712,514
RAC: 0
|
And the last two (141941*2^4299438-1 and 353159*2^4331116-1) have fallen also.
Not quite yet. The last tasks for TRPDC have been send out, but about a dozen are still pending. Once those come back, then it's done.
____________
My lucky number is 75898524288+1 |
|
|
|
|
|
Whats happened with the factor 415267*2^n+1? Its still open. |
|
|
