CRUS LLRnet server discussion - Page 12

mdettweiler · 2008-01-23, 19:01

Okay, I've finally gotten the Perl scripts for manipulating LLRnet results files put together! I was able to generate normal lresults files from the LLRnet results file Carlos posted, so now the Riesel Base 16 range has been posted as complete.

Also, I verified that the 30 doublechecks we ran did come out OK!

Carlos, is the Riesel LLRnet server shut down right now? If it's all cleared out and ready, could you please send me or post here the results file for Karsten's doublechecks that we ran? They should be all prime, so I'd like to use it to test my Perl scripts to see how they handle primes in a results file.

I can then reserve a range on the Riesel Base 16 drive for you to insert into the Riesel server.

Also, am I all clear to post the "official" LLRnet announcement now? Or should I wait for your power supply upgrade?

Stay tuned for some cool stats on the Riesel Base 16 LLRnet run just completed!

em99010pepe · 2008-01-23, 19:22

I'm out to the theater. I'll be away for 4 hours...

mdettweiler · 2008-01-23, 19:49

Okay, here's the stats for the Riesel Base 16 range 87.5K-90K! What do you all think?

Code:

This run of LLRnet tested a total of 808 numbers.
Those 808 numbers were divided up among various users as follows:

Anonymous:   333 numbers tested, 41% of total
tnerual:     167 numbers tested, 20% of total
kar_bon:     126 numbers tested, 15% of total
em99010pepe: 109 numbers tested, 13% of total
j3c2:        73 numbers tested,  9% of total

Primes found: none

I guess I probably could format it a little better--maybe sort them into proper columns with headers and all that--but it's a good start.

mdettweiler · 2008-01-23, 20:34

Carlos, here's some numbers to feed into the Riesel server.

(These are actual production files, not doublechecks.)

BTW, did you remove k=16390 from the Sierp. server yet?

mdettweiler · 2008-01-23, 23:58

Gary has started sieving on a 3rd Drive, which should be all ready by the end of this week at the earliest (probably sometime next week, depending on how many cores he puts on it). The 3rd Drive will be for Base 6 Sierpinski. We'll want to have an LLRnet server set up for Sierp. Base 6--however, I was thinking that we could get it started with the three Sierp. Base 6 k's that already have sieved files available.

Carlos, could you please start up an LLRnet server for Base 6 Sierp (I've got a file attached to this message that you can feed into it)? It doesn't matter to me which port you put it on.

After this LLRnet server is up and running, we won't add any more for the time being--three servers should be plenty of variety for users to pick from, and we'll have all the team drives covered. All three servers will have plenty of work to last them a while, so the only maintenance that will need to be done is removing k's that have had a prime found.

em99010pepe · 2008-01-24, 00:24

Karsten's_results.zip

Riesel Server up and running on port 444.

mdettweiler · 2008-01-24, 00:26

Quote:

Originally Posted by em99010pepe

Karsten's_results.zip

Riesel Server up and running on port 444.

Thanks!

em99010pepe · 2008-01-24, 00:46

Quote:

Originally Posted by Anonymous

Gary has started sieving on a 3rd Drive, which should be all ready by the end of this week at the earliest (probably sometime next week, depending on how many cores he puts on it). The 3rd Drive will be for Base 6 Sierpinski. We'll want to have an LLRnet server set up for Sierp. Base 6--however, I was thinking that we could get it started with the three Sierp. Base 6 k's that already have sieved files available.

Carlos, could you please start up an LLRnet server for Base 6 Sierp (I've got a file attached to this message that you can feed into it)? It doesn't matter to me which port you put it on.

After this LLRnet server is up and running, we won't add any more for the time being--three servers should be plenty of variety for users to pick from, and we'll have all the team drives covered. All three servers will have plenty of work to last them a while, so the only maintenance that will need to be done is removing k's that have had a prime found.

Sorry, I'll only run two servers...three is too much....

mdettweiler · 2008-01-24, 00:49

Well, I checked the results for Karsten's doublechecks--they're all prime, like they were supposed to be. I did uncover a bug in my Perl program that turns the LLRnet results files into CSV files for sorting by n: It wouldn't output the time if a prime was found. I've now fixed that, and it works fine for primes and composites now.

mdettweiler · 2008-01-24, 00:50

Quote:

Originally Posted by em99010pepe

Sorry, I'll only run two servers...three is too much....

Okay, no problem.

I'll go unreserve the k's. Yeah, three servers would be a bit much, especially for a small project like this one--two is enough.

em99010pepe · 2008-01-24, 00:55

Quote:

Originally Posted by Anonymous

Okay, no problem.

I'll go unreserve the k's. Yeah, three servers would be a bit much, especially for a small project like this one--two is enough.

Next time ask me before reserving more work....Tomorrow night the server will be down due to an upgrade of its power supply. I'll let all know when.

2008-01-23, 19:49	#124
mdettweiler A Sunny Moo Aug 2007 USA (GMT-5) 3·2,083 Posts	Okay, here's the stats for the Riesel Base 16 range 87.5K-90K! What do you all think? Code: This run of LLRnet tested a total of 808 numbers. Those 808 numbers were divided up among various users as follows: Anonymous: 333 numbers tested, 41% of total tnerual: 167 numbers tested, 20% of total kar_bon: 126 numbers tested, 15% of total em99010pepe: 109 numbers tested, 13% of total j3c2: 73 numbers tested, 9% of total Primes found: none I guess I probably could format it a little better--maybe sort them into proper columns with headers and all that--but it's a good start. Last fiddled with by mdettweiler on 2008-01-23 at 20:02

2008-01-23, 23:58	#126
mdettweiler A Sunny Moo Aug 2007 USA (GMT-5) 3×2,083 Posts	Gary has started sieving on a 3rd Drive, which should be all ready by the end of this week at the earliest (probably sometime next week, depending on how many cores he puts on it). The 3rd Drive will be for Base 6 Sierpinski. We'll want to have an LLRnet server set up for Sierp. Base 6--however, I was thinking that we could get it started with the three Sierp. Base 6 k's that already have sieved files available. Carlos, could you please start up an LLRnet server for Base 6 Sierp (I've got a file attached to this message that you can feed into it)? It doesn't matter to me which port you put it on. After this LLRnet server is up and running, we won't add any more for the time being--three servers should be plenty of variety for users to pick from, and we'll have all the team drives covered. All three servers will have plenty of work to last them a while, so the only maintenance that will need to be done is removing k's that have had a prime found. Last fiddled with by mdettweiler on 2008-01-26 at 16:39

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
discontinuation of the LLRNet server	opyrt	Prime Sierpinski Project	1	2009-09-07 19:13
LLRnet servers for CRUS	gd_barnes	Conjectures 'R Us	39	2008-07-15 10:26
NPLB LLRnet server discussion	em99010pepe	No Prime Left Behind	229	2008-04-30 19:13
New LLRnet servers discussion	IronBits	Conjectures 'R Us	11	2008-03-20 03:43
llrnet server down?	Molnar	Prime Sierpinski Project	4	2006-10-19 14:32

2008-01-23, 19:01	#122
mdettweiler A Sunny Moo Aug 2007 USA (GMT-5) 1869₁₆ Posts	Okay, I've finally gotten the Perl scripts for manipulating LLRnet results files put together! I was able to generate normal lresults files from the LLRnet results file Carlos posted, so now the Riesel Base 16 range has been posted as complete. Also, I verified that the 30 doublechecks we ran did come out OK! Carlos, is the Riesel LLRnet server shut down right now? If it's all cleared out and ready, could you please send me or post here the results file for Karsten's doublechecks that we ran? They should be all prime, so I'd like to use it to test my Perl scripts to see how they handle primes in a results file. I can then reserve a range on the Riesel Base 16 drive for you to insert into the Riesel server. Also, am I all clear to post the "official" LLRnet announcement now? Or should I wait for your power supply upgrade? Stay tuned for some cool stats on the Riesel Base 16 LLRnet run just completed!

2008-01-23, 19:22	#123
em99010pepe Sep 2004 2·5·283 Posts	I'm out to the theater. I'll be away for 4 hours...

2008-01-24, 00:49	#130
mdettweiler A Sunny Moo Aug 2007 USA (GMT-5) 3·2,083 Posts	Well, I checked the results for Karsten's doublechecks--they're all prime, like they were supposed to be. I did uncover a bug in my Perl program that turns the LLRnet results files into CSV files for sorting by n: It wouldn't output the time if a prime was found. I've now fixed that, and it works fine for primes and composites now.