mersenneforum.org

mersenneforum.org (https://www.mersenneforum.org/index.php)
-   No Prime Left Behind (https://www.mersenneforum.org/forumdisplay.php?f=82)
-   -   Team drive #8 k=1400-2000 n=350K-500K (https://www.mersenneforum.org/showthread.php?t=11245)

gd_barnes 2009-02-11 13:45

Results files have been verified as complete up to n=400K on this drive. Primes were compared to post 1 of this thread, the 8th drive page, Rieselprime.org, and the top-5000 site. No problems were found. :smile:

Ian or Max, when you have a chance, can you coordinate on processing the results for n=400K-440K to me? Thanks.


Gary

mdettweiler 2009-02-11 14:54

[quote=gd_barnes;162457]Results files have been verified as complete up to n=400K on this drive. Primes were compared to post 1 of this thread, the 8th drive page, Rieselprime.org, and the top-5000 site. No problems were found. :smile:

Ian or Max, when you have a chance, can you coordinate on processing the results for n=400K-440K to me? Thanks.


Gary[/quote]
Okay, sure. Though, originally I was planning to wait until 450K to do it--I generally try to wait until I can do an entire contiguous chunk before I process results, unless that chunk is just waaaay to big to do that with. And considering as how there are only two <450K k/n pairs remaining in the server, that shouldn't be too long to wait. :smile:

gd_barnes 2009-02-15 12:11

[quote=mdettweiler;162468]Okay, sure. Though, originally I was planning to wait until 450K to do it--I generally try to wait until I can do an entire contiguous chunk before I process results, unless that chunk is just waaaay to big to do that with. And considering as how there are only two <450K k/n pairs remaining in the server, that shouldn't be too long to wait. :smile:[/quote]


Passed 450K quite a while ago.

mdettweiler 2009-02-15 14:39

[quote=gd_barnes;162869]Passed 450K quite a while ago.[/quote]
Okay, I'll do the results sometime today. :smile:

mdettweiler 2009-02-15 14:59

[quote=gd_barnes;162869]Passed 450K quite a while ago.[/quote]

[quote=mdettweiler;162884]Okay, I'll do the results sometime today. :smile:[/quote]
Uh...actually, I just checked the knpairs.txt on port 8000 just now and this is what the top of the file looks like:[code]
10000000000000:M:1:2:258
1579 448383
1689 448383
1647 458484
1473 458487
1447 458493
1419 458497
1875 458499
1605 458500
1965 458500
1609 458507
1751 458510
1787 458512
1995 458512
1937 458518[/code]As you can see, there are still two results <450K outstanding. No big deal, though; I'll see about cleaning them out on the server in a moment. After that they should disappear after the next time the server prunes knpairs.txt and I can then process the results. :smile:

mdettweiler 2009-02-16 05:17

LLRnet IB8000 has completed 400K-450K; lresults emailed to Gary. :smile:

Karsten, even though I had to manually fill in two results from this range as described in the post above this one, since I submitted them to the server after finishing them, they should be in the 2/16 results file as normally. Thus, everything should be all in place with no further fixes/fill-ins necessary. :smile:

kar_bon 2009-02-16 09:59

[QUOTE=mdettweiler;162961]Karsten, even though I had to manually fill in two results from this range as described in the post above this one, since I submitted them to the server after finishing them, they should be in the 2/16 results file as normally.[/QUOTE]

[b]They are not![/b]

gd_barnes 2009-02-17 11:48

[quote=mdettweiler;162468]Okay, sure. Though, originally I was planning to wait until 450K to do it--I generally try to wait until I can do an entire contiguous chunk before I process results, unless that chunk is just waaaay to big to do that with. And considering as how there are only two <450K k/n pairs remaining in the server, that shouldn't be too long to wait. :smile:[/quote]

[quote=mdettweiler;162892]Uh...actually, I just checked the knpairs.txt on port 8000 just now and this is what the top of the file looks like:[code]
10000000000000:M:1:2:258
1579 448383
1689 448383
1647 458484
1473 458487
1447 458493
1419 458497
1875 458499
1605 458500
1965 458500
1609 458507
1751 458510
1787 458512
1995 458512
1937 458518[/code]As you can see, there are still two results <450K outstanding. No big deal, though; I'll see about cleaning them out on the server in a moment. After that they should disappear after the next time the server prunes knpairs.txt and I can then process the results. :smile:[/quote]


Max,

I'm confused. First, you said there are 2 pairs remaining to be processed. I wait 4 calendar days and state that we passed n=450K because I know that the JobMaxtime is 3 days so that those 2 pairs must have been processed. You then state that there are 2 pairs remaining to be processed again. What gives? Did someone cache those and never process them a second time? That seems unlikely. Were they 2 different k/n pairs then the first time?

Can you research why the results didn't show up in Feb. 16th like Karsten said? Did you submit them after midnight server time such that they will show up in Feb. 17th?

One other thing: Let's not wait until huge contiguous k-ranges like this are complete. Let's do n=20K ranges in the future like we did for the 1st/2nd/3rd drives. Actually, straggling pairs for n>440K is why I requested n=400K-440K to begin with. n=440K-450K could have waited a week or even more if we just did the lower n-range first. Then you wouldn't have had to go through all that hassle on those 2 pairs.

One final thing: I had a reason that I had separated the port 8000 n-ranges into n=20K ranges in the 1st post of this thread. Oh well, I'm not going to do it again. I'll just deal with it.


Gary

gd_barnes 2009-02-17 12:06

Reserving n=465K-475K for port 8000.

kar_bon 2009-02-17 12:27

why got we automated LLR-test?

because all will done automatically!

if the server will wait 3 days before delivering not submitted pairs again, so wait these 3 days!

no need to do them manually!!!!

(please read the thread "NPLB Database" too)

gd_barnes 2009-02-18 18:38

[quote=kar_bon;163118]why got we automated LLR-test?

because all will done automatically!

if the server will wait 3 days before delivering not submitted pairs again, so wait these 3 days!

no need to do them manually!!!!

(please read the thread "NPLB Database" too)[/quote]


I must admit I'm confused by this whole situation too Karsten.

I asked for the results for n=400K-440K because I knew that there were still straggling pairs for n=440K-450K. Max then suggested that we wait for n=400K-450K because only 2 pairs were left so I wait patiently. 4 days later with no word, I requested n=400K-450K since it should be done by then. Max then says there are still 2 pairs left and that they had to be done manually to complete the n-range. This was not my intent at all.

Max, if I may request: Please process the results to me in n=20K ranges after all pairs have been processed automatically. There's no reason to manually process them.

Also, did the 2 pairs that you processed to the server show up in our results on the server yet?


Gary


All times are UTC. The time now is 18:51.

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2022, Jelsoft Enterprises Ltd.