mersenneforum.org

mersenneforum.org (https://www.mersenneforum.org/index.php)
-   No Prime Left Behind (https://www.mersenneforum.org/forumdisplay.php?f=82)
-   -   Individual-k reservations k=300-400 n=1M-3M (https://www.mersenneforum.org/showthread.php?t=13188)

gd_barnes 2010-03-21 03:23

Individual-k reservations k=300-400 n=1M-3M
 
k=300-350 for n=2M-3M is loaded into PRPnet port 1468.
Individual-k reservations are available for k=351-400 n=2M-3M.
.
Well sieved file(s) are available from PrimeGrid. If you need one and don't know where to get it from there, let us know with your reservation and we will send it to you. Please plan to complete any reservations within ~6 months. A status update ~once/month stating where you're at would also be helpful.

As before, please post reservations, statuses, completions, and results in this thread (or Email results to me) and post primes in the primes thread. Partial n-ranges can be reserved. Just state what k and n-range you would like. Please make n-ranges the lowest available.

Let's find some BIG primes! :smile:

reservations and search ranges:
[code]
k n-range reserved by status # primes
301 1M-2M unconnected complete 3
301 2M-3M PRPNet (G1468)
303 1M-2M unconnected complete 0
303 2M-3M PRPNet (G1468) in progress 2
305 1M-2M unconnected complete 0
305 2M-3M PRPNet (G1468) in progress 1
307 1M-1.1M firejuggler complete 0
307 1.1M-2M Lennart complete 0
307 2M-3M PRPNet (G1468) in progress 1
309 1M-1.3M gamer007 complete 0
309 1.3M-2M Lennart complete 1
309 2M-3M PRPNet (G1468)
311 1M-1.6M Cruelty complete 1
311 1.6M-2M unconnected complete 0
311 2M-3M PRPNet (G1468) in progress 1
313 1M-1.2M gamer007 complete 1
313 1.2M-1.3M gd_barnes complete 0
313 1.3M-2M Lennart complete 2
313 2M-3M PRPNet (G1468)
315 1M-2M LLRnet (G3500) complete 5
315 2M-3M PRPNet (G1468)
317 1M-1.2M gamer007 complete 1
317 1.2M-1.3M gd_barnes complete 0
317 1.3M-2M PRPnet (G1468) complete 1
317 2M-3M PRPNet (G1468)
319 1M-1.3M gamer007 complete 0
319 1.3M-2M MyDogBuster complete 3
319 2M-3M PRPNet (G1468) in progress 2
321 1M-2M LLRnet (G3500) complete 2
321 2M-3M PRPNet (G1468) in progress 2
323 1M-1.1M MrOzzy complete 0
323 1.1M-2M Lennart complete 0
323 2M-3M PRPNet (G1468)
325 1M-1.1M nuggetprime PRPnet complete 0
325 1.1M-1.3M gd_barnes complete 0
325 1.3M-2M MyDogBuster complete 0
325 2M-3M PRPNet (G1468) in progress 3
327 1M-2M LLRnet (G3500) complete 1
327 2M-3M PRPNet (G1468)
329 1M-1.3M gd_barnes complete 0
329 1.3M-2M PRPnet (G1468) complete 0
329 2M-3M PRPNet (G1468)
331 1M-1.3M gd_barnes complete 0
331 1.3M-2M kar_bon complete 0
331 2M-3M PRPNet (G1468)
333 1M-2M LLRnet (G3500) complete 4
333 2M-3M PRPNet (G1468) in progress 2
335 1M-1.3M gd_barnes complete 0
335 1.3M-1.366M MyDogBuster complete 0
335 1.366M-2M PRPnet (G1468) complete 3
335 2M-3M PRPNet (G1468) in progress 2
337 1M-2M Cruelty complete 0
337 2M-? Cruelty in progress, n=7.76M
339 1M-2M LLRnet (G3500) complete 3
339 2M-3M PRPNet (G1468) in progress 1
341 1M-1.3M kar_bon complete 1
341 1.3M-2M PRPnet (G1468) complete 1
341 2M-3M PRPNet (G1468)
343 1M-2M Lennart complete 1
343 2M-3M PRPNet (G1468)
345 1M-2M LLRnet (G3500) complete 3
345 2M-3M PRPNet (G1468) in progress 1
347 1M-1.61M LLRnet (G3500) complete 1
347 1.61M-2M PRPnet (G1468) complete 1
347 2M-3M PRPNet (G1468) in progress 1
349 1M-1.5M Mini-Geek complete 2
349 1.5M-2M kar_bon complete 0
349 2M-3M PRPNet (G1468)
351 1M-1.61M LLRnet (G3500) complete 0
351 1.61M-2M PRPnet (G1468) complete 1
353 1M-1.5M Mini-Geek complete 1
353 1.5M-2M MyDogBuster complete 0
355 1M-1.61M LLRnet (G3500) complete 1
355 1.61M-2M PRPnet (G1468) complete 1
357 1M-1.02M Batalov complete 1
357 1.02M-1.61M LLRnet (G3500) complete 2
357 1.61M-2M PRPnet (G1468) complete 3
359 1M-1.61M LLRnet (G3500) complete 0
359 1.61M-2M PRPnet (G1468) complete 0
361 1M-1.61M LLRnet (G3500) complete 3
361 1.61M-2M PRPnet (G1468) complete 0
363 1M-1.3M MyDogBuster complete 0
363 1.3M-2M PRPnet (G1468) complete 2
365 1M-1.61M LLRnet (G3500) complete 0
365 1.61M-2M PRPnet (G1468) complete 0
367 1M-1.61M LLRnet (G3500) complete 1
367 1.61M-2M PRPnet (G1468) complete 0
369 1M-1.61M LLRnet (G3500) complete 0
369 1.61M-2M PRPnet (G1468) complete 1
371 1M-1.61M LLRnet (G3500) complete 3
371 1.61M-2M PRPnet (G1468) complete 1
373 1M-2M MyDogBuster complete 1
375 1M-1.02M Batalov complete 0
375 1.02M-1.61M LLRnet (G3500) complete 4
375 1.61M-2M PRPnet (G1468) complete 0
377 1M-1.61M LLRnet (G3500) complete 1
377 1.61M-2M PRPnet (G1468) complete 0
379 1M-1.61M LLRnet (G3500) complete 1
379 1.61M-2M PRPnet (G1468) complete 2
381 1M-1.61M LLRnet (G3500) complete 1
381 1.61M-2M PRPnet (G1468) complete 2
383 1M-1.61M LLRnet (G3500) complete 1
383 1.61M-2M PRPnet (G1468) complete 2
385 1M-1.61M LLRnet (G3500) complete 1
385 1.61M-2M PRPnet (G1468) complete 0
387 1M-1.61M LLRnet (G3500) complete 0
387 1.61M-2M PRPnet (G1468) complete 3
389 1M-2M MyDogBuster complete 0
391 1M-1.3M MyDogBuster complete 1
391 1.3M-2M PRPnet (G1468) complete 4
393 1M-1.3M MyDogBuster complete 2
393 1.3M-2M PRPnet (G1468) complete 1
395 1M-2M MyDogBuster complete 2
397 1M-2M MyDogBuster complete 0
399 1M-1.3M MyDogBuster complete 1
399 1.3M-2M PRPnet (G1468) complete 4
[/code]primes:
[code]
prime found by 115 total
335*2^2708958-1 ototero
333*2^2591874-1 MiHost
303*2^2562423-1 MiHost
311*2^2544778-1 MiHost
319*2^2498685-1 MiHost
321*2^2496594-1 ototero
325*2^2473267-1 MiHost
303*2^2417452-1 ototero
333*2^2412735-1 MiHost
321*2^2378535-1 MiHost
305*2^2358854-1 MiHost
335*2^2338972-1 ototero
339*2^2336421-1 odicin
345*2^2313720-1 MiHost
347*2^2245598-1 odicin
325*2^2223243-1 ototero
307*2^2137553-1 ototero
319*2^2130729-1 composite
325*2^2024035-1 composite
317*2^1991592-1 Lennart
381*2^1974841-1 Lennart
391*2^1949159-1 odicin
351*2^1947281-1 Lennart
357*2^1934704-1 Lennart
335*2^1917610-1 Lennart
379*2^1909097-1 Lennart
391*2^1886863-1 Lennart
357*2^1871600-1 odicin
315*2^1869119-1 ototero
347*2^1861974-1 odicin
333*2^1853115-1 kar_bon
399*2^1839019-1 Lennart
379*2^1837291-1 Lennart
301*2^1801207-1 unconnected
319*2^1797261-1 MyDogBuster
383*2^1794636-1 Lennart
387*2^1793857-1 odicin
391*2^1780155-1 Lennart
399*2^1764851-1 Lennart
335*2^1762548-1 Lennart
357*2^1756764-1 odicin
355*2^1752713-1 odicin
345*2^1742652-1 kar_bon
315*2^1741334-1 kar_bon
319*2^1740047-1 MyDogBuster
387*2^1732185-1 Lennart
371*2^1717250-1 Mini-Geek
383*2^1716780-1 odicin
387*2^1709440-1 Mini-Geek
313*2^1703119-1 Lennart
395*2^1690690-1 MyDogBuster
399*2^1659001-1 MyDogBuster
309*2^1647947-1 Bibi
381*2^1646029-1 Lennart
315*2^1639432-1 Flatlander
369*2^1635299-1 Lennart
395*2^1611672-1 MyDogBuster
335*2^1597932-1 Mini-Geek
391*2^1596805-1 Kracker
393*2^1583890-1 Mini-Geek
311*2^1581686-1 Cruelty
373*2^1575751-1 MyDogBuster
363*2^1547344-1 kracker
315*2^1539539-1 Flatlander
333*2^1537644-1 Flatlander
347*2^1529964-1 ototero
375*2^1518534-1 ototero
301*2^1514873-1 unconnected
363*2^1513706-1 MyDogBuster
341*2^1484130-1 MyDogBuster
377*2^1483586-1 MyDogBuster
333*2^1474766-1 Flatlander
327*2^1473201-1 Flatlander
357*2^1473125-1 MyDogBuster
313*2^1456431-1 Lennart
399*2^1418376-1 MyDogBuster
371*2^1417702-1 MyDogBuster
339*2^1410789-1 kar_bon
321*2^1371846-1 kar_bon
321*2^1363671-1 kar_bon
367*2^1352793-1 kar_bon
361*2^1346489-1 MyDogBuster
319*2^1322059-1 MyDogBuster
375*2^1319127-1 kar_bon
379*2^1302991-1 MyDogBuster
357*2^1301704-1 MyDogBuster
393*2^1297402-1 MyDogBuster
399*2^1293056-1 MyDogBuster
349*2^1278551-1 Mini-Geek
375*2^1275345-1 MyDogBuster
391*2^1247959-1 MyDogBuster
349*2^1228715-1 Mini-Geek
333*2^1209174-1 kar_bon
371*2^1180806-1 MyDogBuster
345*2^1175842-1 kar_bon
339*2^1161347-1 kar_bon
301*2^1157561-1 unconnected
341*2^1129218-1 kar_bon
385*2^1126083-1 MyDogBuster
355*2^1125573-1 kar_bon
315*2^1116547-1 kar_bon
353*2^1114356-1 Mini-Geek
313*2^1137387-1 gamer007
371*2^1116274-1 ototero
339*2^1110323-1 kar_bon
393*2^1095688-1 MyDogBuster
383*2^1091332-1 ototero
381*2^1070022-1 MyDogBuster
375*2^1066662-1 kar_bon
361*2^1043099-1 MyDogBuster
361*2^1033973-1 MyDogBuster
317*2^1032378-1 gamer007
343*2^1020791-1 Lennart
357*2^1019004-1 Batalov
345*2^1000251-1 kar_bon
[/code]Gary

MyDogBuster 2010-03-21 05:47

Reserving k = 395, 397 and 399 to whatever is available. What is the cutoff at this level of sieve anyway?

Gary, I need a copy of the sieve file so I can disperse ranges.

unconnected 2010-03-21 06:00

I would like to reserve 303 and 311 (from 1.6M). Please send me the files.

gd_barnes 2010-03-21 06:08

[quote=MyDogBuster;209077]Reserving k = 395, 397 and 399 to whatever is available. What is the cutoff at this level of sieve anyway?

Gary, I need a copy of the sieve file so I can disperse ranges.[/quote]

Yeah, I was going to send the file to you and Max here shortly.

There is no cutoff. Search to whatever you want. We'll do some more sieving but the optimum depth for the entire range is not much higher than what it is already sieved to; perhaps P=175T. I wasn't going to hold the drive up any longer because it's already a far deeper sieve than anyone would search to if they were just doing a few k's on their own and the drive is likely to last 3-5 years anyway.

Unconnected and Ian, n=1M-2M is a tremendous amount of work. Are you sure you want to reserve the entire range on your reserved k's? It will likely take at least 10 times longer to search n=1M-2M than n=600K-1M.


Gary

MyDogBuster 2010-03-21 06:21

I'll reserve those 3 k's (395, 397 & 399) to n=1.3M for now.

gd_barnes 2010-03-21 07:49

Dmitry,

I Emailed you the 2 sieve files up to n=2M. Let me know if you want a smaller reservation.


Gary

unconnected 2010-03-21 08:47

Thanks, Gary!
I'll start testing and will see how it's going.

mdettweiler 2010-03-29 03:08

I was going to post this earlier when this thread was first started but I forgot: as before with the n=600K-1M individual-k drive, if anyone would like to coordinate their resources by means of an automated server like we have for our team drives, I'll be glad to host one for them. I've found that a server comes in quite handy when dealing with multiple cores/machines, especially if some of them are not easily accessible. I can do both LLRnet and PRPnet servers (for PRPnet, I'm currently using v2.4.6 for all servers, but since 3.2.3 sounds like it will finally be able to work stably on Linux machines such as our server, I'd be willing to use that if requested); results would be posted to a downloadable URL like with the NPLB/CRUS public servers, and I'd process the raw results into the standard sorted format that Gary needs them in.

Note that this isn't limited just to reservations from this individual-k drive; I'm willing to host pretty much anything that LLRnet or PRPnet will support, since these servers aren't that hard to run once you've got the infrastructure already set up. :smile: One possible case this may be useful for is if, say, you have all your clients configured for PRPnet and still want to use it even though we're moving away from it on NPLB public servers now; one person in particular emailed me regarding such a conundrum and I'm thinking of him here. :smile:

Cruelty 2010-03-31 16:37

Status report
 
k=337 tested till n=2.39M - I'm still working on it.

Mini-Geek 2010-04-07 13:17

I'll take k=349 and k=353 from 1M to 1.5M. (yep, the same ones I searched from 600K-850K...maybe they're due for primes now :grin:)

Edit: Files sent

Mini-Geek 2010-04-07 17:59

[quote=Mini-Geek;210843]I'll take k=349 and k=353 from 1M to 1.5M.

Edit: Files sent[/quote]
Files received and work has begun. I don't really want to calculate an ETA, (might give me second thoughts about this :razz:) but for the most part it will have the full attention of my quad, and I'll report on it periodically.
By a remarkable coincidence, there are exactly as many candidates for each k in the range I reserved! Each has 14561 candidates in n=1M-1.5M. I originally (for n=600K+) picked them because they had similar weights, but this is still a very odd coincidence. (over 1M-2M, 353 has 120 more candidates than 349, which is actually opposite from what you'd expect from their Nash weights: 349's 1463 vs 353's 1416)

kar_bon 2010-04-08 21:47

I've updated the 'Individual k'-page [url=www.rieselprime.de/NPLB/Efforts/Individual-k.htm]here[/url].

firejuggler 2010-04-26 12:15

so , can I get k =307 from 1M to 1.25M?

firejuggler 2010-04-26 19:59

ok, after a bit of testing, i'm going to reduce the range... 1M to 1.1M

Edited MyDog: I will send the file after I receive an email address via PM

Edit: File sent

gd_barnes 2010-04-27 07:24

Ian, thanks for sending the file to him.

Firejuggler, welcome to NPLB! I've been on a business trip so sorry to have not responded sooner. If you have any questions, let us know. One of the 4 admins is bound to respond fairly quickly. :-)

firejuggler 2010-04-27 10:04

no problem.
To split a file, you just have to divide roughly the line by 2 and paste the first line, right? ( and set aa different cpu for each LLR), right?
ie : i have a file
starting with
140000000000000:M:1:2:258
307 1000037
307 1000201
and the other with

140000000000000:M:1:2:258
307 1048321
307 1048333

kar_bon 2010-04-27 10:45

Hi firejuggler,

yep, that's correct.

The header-line is needed in every candidates file at the top to say LLR what work to do.

If you got more than one core, you can also try to work with a LLRnet-server-client. For a download and more information see [url=http://www.mersenneforum.org/showthread.php?t=13165]here[/url].

firejuggler 2010-04-27 16:16

hi.
crunching started...
I do not manage to use the server-client one...
(or more exactly , configure it)
so I use 2 LLR , one on each core

Cruelty 2010-04-30 18:36

Status report
 
k=337 tested till n=2.45M - I'm still working on it.

Mini-Geek 2010-05-09 14:01

349 and 353 are at n=1.11M. No primes.

gd_barnes 2010-05-14 22:52

Hi all,

What does everyone think about another 6k LLRnet mini drive here? I'm thinking the same 6 k's as before...all divisible by 3...those being k=315, 321, 327, 333, 339, & 345. Initially, we would start with the range of n=1M-1.3M and then decide what we want to do after that.

Max, assuming that we decide to do this, I'll leave it up to you what port we want to use. I seem to remember you mentioning port 2500 in an Email but wasn't sure if that was set in stone. If the interest is there, we could also set up a PRPnet server for part of the range.

If anyone has any different ideas about it, let us know. The drive will be informal like before with no separate thread.


Gary

kar_bon 2010-05-14 22:58

[QUOTE=gd_barnes;215048]What does everyone thing about another 6k LLRnet mini drive here? I'm thinking the same 6 k's as before...all divisible by 3...those being k=315, 321, 327, 333, 339, & 345.[/QUOTE]

Good idea and I was waiting for such effort. It's easier to maintain and to process.
And I want to get back my 13-out-of-20-biggest-LLRnet-NPLB-primes-place! :grin:

mdettweiler 2010-05-15 04:12

[quote=gd_barnes;215048]Max, assuming that we decide to do this, I'll leave it up to you what port we want to use. I seem to remember you mentioning port 2500 in an Email but wasn't sure if that was set in stone. If the interest is there, we could also set up a PRPnet server for part of the range.[/quote]
The port I previously mentioned to Dave for this was 3500--I've reserved the 2x00 space for doublecheck servers in the future (to avoid confusion with standard servers). I already detailed my numbering convention to Gary by email before, but for everyone else's sake, here it is:

First-pass NPLB servers: ports x000 and x500 (the latter beginning at 3500)
Second-pass NPLB servers: ports 2x00
CRUS servers: ports 1x00

(Note that ports <1024 are a pain to use under Linux, so those are off-limits for servers.)

Right now the only remaining unused x000 port is 2000, so I emailed Dave a while back asking him to open up x500 so we'd have more room there. Even though port 2000 is available, to avoid confusion with future and past doublecheck servers in the 2x00 range, I'd rather avoid using 2000 for firstpass servers unless we completely run out of other options; therefore the new server will be on port 3500. Any more servers in the future will go on 4500, 5500, etc.

Meanwhile, though, definitely agreed on doing another mini-drive. (Duh, Gary and I already discussed this privately so of course I'm agreed on it. :rolleyes:) I might chip in at some point once it's running--as I recall I netted a couple of big juicy primes off the last such mini-drive and I wouldn't mind repeating that, this time for some megabits. :smile:

MyDogBuster 2010-05-18 02:15

395, 397 and 399 at n=1.1M, continuing

gd_barnes 2010-05-18 11:52

To make it official: Reserving k=315/321/327/333/339/345 for n=1M-1.3M for port 3500 for another 6k mini drive.

Max,

I have port 3500 set up, loaded up and ready to go for the next 6k drive but have not yet started it. My router congif has also been tweaked for the new server. For now, I've only loaded n=1M-1.1M into the server.

I can't remember what you did behind-the-scenes for the previous 6k drives. Should it be added to the regular "LLRnet servers for NPLB" web page? If so, of course we'll need to add it to the 1st post of that thread there.

I wanted to hold off starting it until you confirm everything that needs to be done but most importantly, can you please make sure that AMDave has all of the behind-the-scenes stuff set up for the stats DB and that he has had plenty of time to check things in that regard?

After doing what you need to, feel free to let 'er rip. You should just be able to hit enter on the L3500 command prompt server tab.

Thanks! :smile:


Gary

mdettweiler 2010-05-18 14:13

[quote=gd_barnes;215294]To make it official: Reserving k=315/321/327/333/339/345 for n=1M-1.3M for port 3500 for another 6k mini drive.

Max,

I have port 3500 set up, loaded up and ready to go for the next 6k drive but have not yet started it. My router congif has also been tweaked for the new server. For now, I've only loaded n=1M-1.1M into the server.

I can't remember what you did behind-the-scenes for the previous 6k drives. Should it be added to the regular "LLRnet servers for NPLB" web page? If so, of course we'll need to add it to the 1st post of that thread there.

I wanted to hold off starting it until you confirm everything that needs to be done but most importantly, can you please make sure that AMDave has all of the behind-the-scenes stuff set up for the stats DB and that he has had plenty of time to check things in that regard?

After doing what you need to, feel free to let 'er rip. You should just be able to hit enter on the L3500 command prompt server tab.

Thanks! :smile:


Gary[/quote]
Dave should be all set on his end, so the only further things to do are:
-Add the server to the script that copies off its results files daily
-Add the server to my script that generates the [URL]http://www.noprimeleftbehind.net/llrnet/[/URL] page
-Add the server to the "LLRnet servers for NPLB" thread

I'll do both shortly.

Edit: all done!

firejuggler 2010-05-19 00:34

almost done with assignement (307 1M-1.1M, no prime so far)
i'll take 363, for the same range (hopefully, it will go faster)

gd_barnes 2010-05-19 06:57

To make it official:

The 6k mini drive for k=315/321/327/333/339/345 for n=1M-1.3M has now started. The LLRnet server info. is:

server = "www.noprimeleftbehind.net"
port = 3500

gd_barnes 2010-05-19 07:02

[quote=firejuggler;215348]almost done with assignement (307 1M-1.1M, no prime so far)
i'll take 363, for the same range (hopefully, it will go faster)[/quote]

Ian,

Can you Email Firejuggler k=363 for n=1M-1.1M? I think you have his Email address.

If you can copy me in on that, I'll be able to Email them to him in the future.

Firejuggler,

k=363 should take 4-5 times longer to test than k=307. k=363 has 6173 pairs to test for n=1M-1.1M while k=307 only has 1343 pairs. On the bright side, you'll have a lot better chance of a prime. :-)

If you look at the 1st post, you'll see a list of "weights" for k's that have not been started yet. (I've now removed your k=363.) This gives a relative idea of how many pairs each k will have. The higher it is, the more pairs it will have. The average is about 1600. k=363 has a weight of >3100; one of the highest of all k=300-400.


Gary

Edit: File sent

firejuggler 2010-05-19 16:14

ok, so it is the opposite of what I though... I though that the higher the Nash weight the less pair there was to test
well, I'll do it, still.. but it wil be long...

kar_bon 2010-05-19 21:22

Correct, the higher the Nash weight, the more candidates to test.

See [url=http://www.mersenneforum.org/showpost.php?p=211077&postcount=19]here[/url] for more details.

firejuggler 2010-05-19 21:23

1 Attachment(s)
here is 307-1M-1.1M result, no prime
[ATTACH]5222[/ATTACH]

Cruelty 2010-05-31 16:33

Status report
 
k=337 tested till n=2.5M - I'm still working on it.

Mini-Geek 2010-06-01 16:25

k=349 and 353 are at n=1.2M.

kar_bon 2010-06-21 10:49

I've updated the k=300-2000 page on rieselprime.de with the data from here.

I also included a new page for the [url=http://www.rieselprime.de/NPLB/Drives/NPLB_MiniDrive4.htm]Mini Drive #4[/url] (315, 321, 327, 333, 339, 345) and updated the Individual-page.

unconnected 2010-06-26 05:14

k=311 done [1600k-2000k]
k=303 done [1000k-1500k]
k=303 ~75% done [1500k-2000k]

No primes, results are sent via email.
I would like to reserve k=301 and k=305.

gd_barnes 2010-06-27 08:37

[quote=unconnected;219929]I would like to reserve k=301 and k=305.[/quote]

I have now Emailed you the files for k=301 and 305. Good luck! :smile:

unconnected 2010-06-27 10:50

Thank you. Hope I'll find prime sooner or later.

Cruelty 2010-06-30 18:34

Status report
 
k=337 tested till n=2.56M.

Mini-Geek 2010-07-04 22:09

k=349 and 353 are at n=1.3M.

kar_bon 2010-07-05 22:17

@Gary, Max:

I've checked the results for the MiniDrive #4 (6 k's from 1M to 2M).
I've noticed, that the pair 315*2^1000000-1 was done twice!

This pair was tested with the MiniDrive #1 (same 6 k's from 600k-1M).
The resultfile "results_20090909_0001_GB_nplb_8000.txt" contains this pair.

In MiniDrive #4 it was tested again, because the n-range was/is xxx000-xxx000 instead of xxx001-xxx000.
So resultfile "results_20100519_0001_GB_nplb_3500.txt" contains this pair, too.

Please be sure to upload the range from 001-000 into the server.
Have a look at the sievefile you sent me for k=377: it contains the n-values for n=1000000 [b]and[/b] n=2000000!

PS:
There're 3 k-values with n=1000000 in that sievefile: k=315, 377 and 383.

gd_barnes 2010-07-06 07:00

[quote=kar_bon;220649]@Gary, Max:

I've checked the results for the MiniDrive #4 (6 k's from 1M to 2M).
I've noticed, that the pair 315*2^1000000-1 was done twice!

This pair was tested with the MiniDrive #1 (same 6 k's from 600k-1M).
The resultfile "results_20090909_0001_GB_nplb_8000.txt" contains this pair.

In MiniDrive #4 it was tested again, because the n-range was/is xxx000-xxx000 instead of xxx001-xxx000.
So resultfile "results_20100519_0001_GB_nplb_3500.txt" contains this pair, too.

Please be sure to upload the range from 001-000 into the server.
Have a look at the sievefile you sent me for k=377: it contains the n-values for n=1000000 [B]and[/B] n=2000000!

PS:
There're 3 k-values with n=1000000 in that sievefile: k=315, 377 and 383.[/quote]


Karsten,

Is this really a big deal? I do the 1n overlap on all files that I sieve.


Gary

kar_bon 2010-07-06 09:24

[QUOTE=gd_barnes;220659]Karsten,

Is this really a big deal? I do the 1n overlap on all files that I sieve.


Gary[/QUOTE]

Because it was your demand to all who reserve manual ranges to do the xxx001 to xxx000 range to be sure testing without losing any pair!

Anyway, no problem for me.

gd_barnes 2010-07-06 11:50

[quote=kar_bon;220665]Because it was your demand to all who reserve manual ranges to do the xxx001 to xxx000 range to be sure testing without losing any pair!

Anyway, no problem for me.[/quote]

??

There are 3 reasons why this does not apply here:

1. Because the 2 individual-k drives are separate drives and sieving efforts. The first individual-k drive ended at n=1M and the new one restarted from n=1M. This will not come up for n>1M.

2. The admins send people ranges with this drive and know to use xxx001 to xxx000 when sending files to people for ranges n>1M.

3. We never said anything about it in the 1st post here because of the above 2 reasons. It's only stated on drives where people are pulling subranges out of larger posted files both here and at CRUS. We only post small beginning files here.


You may remember that you made a similar statement about n=600000 for all of the various k=300 to 1001 drives. The main reason that I asked you if it's a problem for you now is that it will happen again at n=1M for other various drives. If it was creating a problem, I/we could remove all tests for n=1M in the future when we are sieving k=400-1001 for n=1M-2M but the files are so large and difficult to manipulate that's it's my preference to either (1) leave them in there (2) be very specific by sieving n=1000001 to 2000000. I guess it is just easier to sieve with -n 1e6 -N 2e6 rather than -n 1000001 -N 2e6.

As for how to handle them in the database when they are loaded: Make them a doublecheck.


Gary

kar_bon 2010-07-06 12:34

And a big reason on my side:

I'm checking [b]all[/b] resultfile from LLRnet/PRPnet for all drives and it's highly automated to save me time instead of checking things by hand.
One script takes the big sievefile and extracts the recommended candidates for a 1k or 10k range beginning with xxx001.
Doing the same only at the beginning of a new sieverange like 1M-2M will pop up errors like pair is not in that automaticly created sieverange, so i have to check why by hand.
That's what I want to avoid.

gd_barnes 2010-07-06 19:58

OK. I'll make a mental note to sieve only xxxx001 to xxxx000 on various drives in the future -IF- NPLB originally sieved the range immediately below it. Afaik, that should be the case on all drives at this point.

MyDogBuster 2010-07-12 05:38

395, 397 and 399 at n=1.2M, continuing

Cruelty 2010-08-01 17:15

Status report
 
k=337 tested till n=2.62M

kar_bon 2010-08-09 02:05

There's about 3 days work left at port 3500 for me!

gd_barnes 2010-08-09 21:20

[quote=kar_bon;224515]There's about 3 days work left at port 3500 for me![/quote]

We went through that quicker than expected. n=1.1M-1.2M will be loaded within the next few hours.

To make it official: Reserving n=1.1M-1.2M for port 3500.

kar_bon 2010-08-10 19:15

[QUOTE=gd_barnes;224657]We went through that quicker than expected. n=1.1M-1.2M will be loaded within the next few hours.
[/QUOTE]

In about 3 hours I'm sending back 356 tested pairs, so I can't reserve this amount again!
Currently 571 pairs untested minus the 356 -> Server is out of work!

gd_barnes 2010-08-10 21:52

[quote=kar_bon;224786]In about 3 hours I'm sending back 356 tested pairs, so I can't reserve this amount again!
Currently 571 pairs untested minus the 356 -> Server is out of work![/quote]

Done. Sorry about the delay.

kar_bon 2010-08-10 21:54

[QUOTE=gd_barnes;224814]Done. Sorry about the delay.[/QUOTE]

Just want to submit and receive more!

MyDogBuster 2010-09-01 03:51

Reserving k's 389, 391 and 393 to n=1.3M

395, 397 and 399 will be finished in a few days to n=1.3M

Cruelty 2010-09-01 22:25

Status report
 
k=337 tested till n=2.67M

MyDogBuster 2010-09-07 18:54

k=395, 397, 399
 
k=395, 397, 399 tested n=1M-1.3M

Nothing found - Results emailed

Mini-Geek 2010-09-07 23:02

k=349 and 353 are at n=1.4M. No further primes to report.
This n=100K section took longer than expected because while I was away, the program I use to run/manage my clients on my computer (PSP Runner by DigiK-oz) stopped working for no apparent reason.

MyDogBuster 2010-09-26 16:53

k=389
 
k=389 tested n=1M-1.3M

Nothing found - Results emailed

Cruelty 2010-09-30 20:03

Status report
 
k=337 tested till n=2.73M

Batalov 2010-10-08 17:26

k=357
 
I took NPLB357-1000-1020.txt and NPLB375-1000-1020.txt for a test of speed.

Batalov 2010-10-09 06:32

1 Attachment(s)
k=357 results, n=1M-1.02M. No further reservation.
(six tests were repeated with LLR, to get a feeling of speed. Same. Both nice programs.)

Batalov 2010-10-15 11:23

1 Attachment(s)
k=375 results, n=1M-1.02M. No further reservation. 0 primes.

Mini-Geek 2010-10-20 03:24

k=349 and 353 are done to n=1.5M. All three primes previously reported. Results in NPLB database already (it uploaded every day, the last one is uploaded and should be imported at the end of the hour or day). I can email a zip of all the results if you want (I saved them all, in both the original PRPnet and the converted LLRnet formats).

mdettweiler 2010-10-20 15:49

[QUOTE=Mini-Geek;233948]k=349 and 353 are done to n=1.5M. All three primes previously reported. Results in NPLB database already (it uploaded every day, the last one is uploaded and should be imported at the end of the hour or day). I can email a zip of all the results if you want (I saved them all, in both the original PRPnet and the converted LLRnet formats).[/QUOTE]
The results in the database should be fine (to clarify: yes, they're imported at the end of the day; only the NPLB public servers are done hourly). I have a handy dandy web page that I can use to pull automatically-sorted results for any range of k and n out of the database, so once you're done with these k's I can process the results very easily.

mdettweiler 2010-10-20 15:54

[QUOTE=mdettweiler;233993]The results in the database should be fine (to clarify: yes, they're imported at the end of the day; only the NPLB public servers are done hourly). I have a handy dandy web page that I can use to pull automatically-sorted results for any range of k and n out of the database, so once you're done with these k's I can process the results very easily.[/QUOTE]
Oops, never mind, I see now that 1.5M [i]was[/i] the end of your reservation. :smile: In that case, I'll get the results processed as soon as I get the chance and email them to both you and Gary.

firejuggler 2010-10-20 18:50

the computer on which i ran 363 crashed. i lost all the data. and no backup. sorry, so i'm releasing it

Cruelty 2010-10-31 14:06

Status report
 
k=337 tested till n=2.79M

gamer007 2010-11-27 06:38

Reserving k=309, 313, 317, 319.

gd_barnes 2010-11-27 06:48

[QUOTE=gamer007;238850]Reserving k=309, 313, 317, 319.[/QUOTE]

Gamer, what n-range of these would you like to reserve? n=1M-2M is available.

gamer007 2010-11-27 07:31

[QUOTE=gd_barnes;238851]Gamer, what n-range of these would you like to reserve? n=1M-2M is available.[/QUOTE]

I'll start with n=1M-1.02M since I already downloaded the ones posted.

MyDogBuster 2010-11-30 06:52

k=391
 
k=391 tested n=1M-1.3M

1 prime previously reported - Results emailed

Cruelty 2010-11-30 22:13

Status report
 
k=337 tested till n=2.85M

MrOzzy 2010-12-07 07:33

taking k=323 n=1m - 1.02m

gd_barnes 2010-12-07 12:34

Gamer is reserving k=309, 313, 317, 319 to n=1.1M.

Max, have you sent him the files yet?

mdettweiler 2010-12-07 19:03

[QUOTE=gd_barnes;240492]Gamer is reserving k=309, 313, 317, 319 to n=1.1M.

Max, have you sent him the files yet?[/QUOTE]
Oh, whoops...I assumed you'd be handling that. I can send him files from the latest RSP2M file (p=1405T) but it will take a day or so since I've only preprocessed (namely: used srfile to convert it to NewPGen format, the most time-consuming step which takes a good 3-4 hours) the RSP1M file that we've been using for the team drives. I'll have to do that for the RSP2M file before I can extract his k's.

kar_bon 2010-12-07 19:11

[QUOTE=mdettweiler;240543]Oh, whoops...I assumed you'd be handling that. I can send him files from the latest RSP2M file (p=1405T) but it will take a day or so since I've only preprocessed (namely: used srfile to convert it to NewPGen format, the most time-consuming step which takes a good 3-4 hours) the RSP1M file that we've been using for the team drives. I'll have to do that for the RSP2M file before I can extract his k's.[/QUOTE]

Which option you use with srfile?

'-G' will create one big file in newpgen-format and sure this will take much more time.

Better use '-g' which will create for every k-value it's own file.
Than copy all files needed in one, delete all header lines (only first remain) and call srfile with '-G' for this file and you got it.

mdettweiler 2010-12-07 19:18

[QUOTE=kar_bon;240546]Which option you use with srfile?

'-G' will create one big file in newpgen-format and sure this will take much more time.

Better use '-g' which will create for every k-value it's own file.
Than copy all files needed in one, delete all header lines (only first remain) and call srfile with '-G' for this file and you got it.[/QUOTE]
That's what I just discovered. The preprocessing finished just a few minutes ago on -g format (which worked perfectly, since for this drive we need separate files for each k). I've emailed the files to you, Gary, and Ian, and will send gamer007 his files shortly.

I'll have to try the trick you suggest next time I need to preprocess the 1M file...I didn't know srfile can go from -g to -G faster than it can go from -a to -G.

kar_bon 2010-12-07 19:27

[QUOTE=mdettweiler;240548]I'll have to try the trick you suggest next time I need to preprocess the 1M file...I didn't know srfile can go from -g to -G faster than it can go from -a to -G.[/QUOTE]

The abcd-format is already sorted by k-values, but differently stored, so srfile only have to find the header of the k-value and can create the newpgen-format easily.

The whole 160M or more candidates with -G take time to sort and write.

mdettweiler 2010-12-07 19:31

[QUOTE=kar_bon;240551]The abcd-format is already sorted by k-values, but differently stored, so srfile only have to find the header of the k-value and can create the newpgen-format easily.

The whole 160M or more candidates with -G take time to sort and write.[/QUOTE]
Yeah, I guessed something like that. I guess what I'm wondering is, is it faster to go from -g to -G, or from -a to -G? Or no difference? (The file needs to be in -G format, i.e. sorted by n primary and k secondary, for the team drives.)

gamer007 2010-12-07 19:31

Thanks guys.

gd_barnes 2010-12-09 04:58

[QUOTE=mdettweiler;240543]Oh, whoops...I assumed you'd be handling that. I can send him files from the latest RSP2M file (p=1405T) but it will take a day or so since I've only preprocessed (namely: used srfile to convert it to NewPGen format, the most time-consuming step which takes a good 3-4 hours) the RSP1M file that we've been using for the team drives. I'll have to do that for the RSP2M file before I can extract his k's.[/QUOTE]

Oh, didn't I send you a PM asking you to do it since I was out of town? I'm just checking to make sure you received the PM.

mdettweiler 2010-12-09 05:20

[QUOTE=gd_barnes;240881]Oh, didn't I send you a PM asking you to do it since I was out of town? I'm just checking to make sure you received the PM.[/QUOTE]
No, I didn't get any PM about that. The last one I have from you is dated "04 Dec 10 07:04 PM" (EST).

(BTW: in the future, you may want to use the "read receipt" feature of PMs. That will put a link on your main PM inbox page which, when clicked, shows you if and when all read-receipted PMs you've sent have been read. I use it all the time...that way I know if someone actually just hasn't checked their PM box yet or if they're ignoring my message. :wink:)

gd_barnes 2010-12-09 15:23

I erroneously sent it to Gamer instead. Gad, the forwarding and responding to of PMs when there is more than one person involved is a pain here.

gd_barnes 2010-12-09 15:28

MrOzzy has reserved k=323 for n=1M-1.4M.

MrOzzy 2010-12-16 06:08

status update for k=323

Testing at n=1.05m, no primes

Note: FFTlen increase from 65536 to 81920 between n=1.051m and 1.052m. Testing now takes 35% longer.

MrOzzy 2010-12-27 06:04

1 Attachment(s)
k=323 finished untill n=1,1m,no primes, results attached.

next report at 1.2m.

Cruelty 2011-01-03 10:37

k=337 tested till n=2.94M

gamer007 2011-01-11 03:12

Now that the rally is done, back to this.
k=309 tested until n=1.05M
k=313 tested until n=1.06M
k=317 tested until n=1.06M
k=319 tested until n=1.05M

nuggetprime 2011-01-14 15:19

Taking k=325 to 1.1M:coffee:
please send candidates split into 8 parts to ilikelinustorvalds AT gmail DOT com
should take me 5-6 months if thats okay

thanks

mdettweiler 2011-01-14 18:30

[QUOTE=nuggetprime;246338]Taking k=325 to 1.1M:coffee:
please send candidates split into 8 parts to ilikelinustorvalds AT gmail DOT com
should take me 5-6 months if thats okay

thanks[/QUOTE]
No problem--5-6 months for a reservation is a somewhat typical timeframe on this drive. :smile:

Regarding splitting the candidates into 8 parts: I can do that, but note that because the tests get bigger as n increases, they would not all finish at the same time. If you'd prefer a more even work distribution, I'd be happy to host a private PRPnet or LLRnet server on this range for you (as I've done for a few others in similar situations).

nuggetprime 2011-01-14 19:53

Yes,please set up a PRPnet server for that!
But I doesn't need it to be private,everybody should be allowed to put his/her cores on it.
As long as it doesn't run dry,I'm fine:smile:

Flatlander 2011-01-14 20:30

NewPGen can split files. I used it recently to split a file for 4x2.66GHz and 4x1.5GHz so that everything finishes together. In theory anyway!

mdettweiler 2011-01-14 20:56

[QUOTE=nuggetprime;246379]Yes,please set up a PRPnet server for that!
But I doesn't need it to be private,everybody should be allowed to put his/her cores on it.
As long as it doesn't run dry,I'm fine:smile:[/QUOTE]
Okay, will do. I'll be in touch with you regarding the server port once I get it set up.

As for it being "private", by that I mean that it would be crunching your personal reservation. Anyone aware of the port # is of course able to contribute to it, but their contributions would effectively go uncredited. Such ranges are classified as "manual" work as far as the stats are concerned, so it would all go in under your name.

As an example of how this has been done in the past, on the (now concluded) k=3010-3200 mini-drive, Gary had for a while a private PRPnet server that he used to distribute work among a couple of his machines. Karsten and I also ran some clients on it for a while, and in Karsten's case even found a few primes. In the end, though, the entire range was imported into the database under Gary's name.

Meanwhile, though, we do have a proper public server for this individual-k drive if anyone else is interested in doing normal, credited contributions--that's LLRnet port 7000. So far there has not been enough demand for an additional PRPnet server on this drive, though if interest is expressed we could of course set one up.

nuggetprime 2011-01-15 11:19

Machine configured to use new private LLRnet server,it's completing pending tasks on port 9000 now.
BTW,why have you added me to team ROLP:razz:?

mdettweiler 2011-01-15 18:53

[QUOTE=nuggetprime;246510]Machine configured to use new private LLRnet server,it's completing pending tasks on port 9000 now.
BTW,why have you added me to team ROLP:razz:?[/QUOTE]
Not to worry, the one ROLP result was from me when I tested the server. Yours will go under whatever team you have configured (presumably Free-DC), or if you're using a pre-4.1.x client they won't be counted in the team stats at all. (That's how I have two results under my name but one under ROLP--I submitted one with 4.0.6 and one with 4.1.3 after I upgraded my clients.)

Note that, of course, since this is a personal server, all the results will in the end be imported into the main database under nuggetprime/Free-DC.

@everyone else: In case you're wondering what the heck I'm talking about here, the 4.1.x clients now support a "team" field for use in calculating PRPnet's internal stats.

nuggetprime 2011-01-15 19:23

thanks!
Just noticed that I'll be able to LLR more than 80 candidates per day ,so it will probably take less than three months to get to 1.1M

MyDogBuster 2011-01-20 02:11

k=393
 
k=393 tested n=1M-1.3M

2 primes previously reported - Results emailed


All times are UTC. The time now is 17:01.

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