LLRnet servers for NPLB - Page 103

MyDogBuster · 2009-07-25, 20:49

Quote:

Also, I agree with what Gary's been saying about running LLRnet servers on his machines. We seem to be running into one problem after another. If everyone is OK with the idea of moving the servers to David's machines, I can package up the knpairs.txt files and send them to David so he can load them into new servers.

I disagree. If the knpairs.txt file for GB4000 is leaving that server then I would like to stake a claim to it and put it on my own server on MY machine.

mdettweiler · 2009-07-25, 20:49

Here's the IP address that I was seeing on the G8000 console output. I can't be sure if this is the same one that pulled down all those k/n pairs since LLRnet doesn't correlate the two in its output, but here it is anyway:

97.222.133.84 (84.sub-97-222-133.myvzw.com)

And its estimated location is:
Country: UNITED STATES (US)
City: Bedminster, NJ

Anybody know of a guy from around the New Jersey area (or within the nearby states) who might even be remotely interested in prime searching or NPLB?

mdettweiler · 2009-07-26, 00:48

I am pleased to announce that I've cleared up the mess on the GB servers, and have sent an email to David requesting that he do the same on the IB servers. As various users discovered earlier, it seems that some unknown person has sucked our servers dry of available work. No big deal; as soon as David's finished clearing things up on his end, we should be all back to normal.

Let's keep those primes coming!

gd_barnes · 2009-07-26, 05:17

OK, everyone. The servers appear to be back to normal. We think that a new person accidentally cached a few too many pairs and dried us out temporarily. Everything looks good now.

Thanks to David and Max for working diligently to get the problem resolved.

Sorry about the problems.

Gary

vaughan · 2009-07-26, 06:24

Can anyone help me with this question please?

How many kb are the NPLB tasks?
What would be the expected d/l kb per day for say an Intel Q6600?

I ask as I'm running out of peak download allowance from my ISP and need a project that doesn't consume heaps (like BOINC Rosetta does).

IronBits · 2009-07-26, 07:24

That's funny about the servers getting sucked dry.

Let me know if it ever happens again so I can fix it for you.

gd_barnes · 2009-07-26, 09:26

Quote:

Originally Posted by vaughan

Can anyone help me with this question please?

How many kb are the NPLB tasks?
What would be the expected d/l kb per day for say an Intel Q6600?

I ask as I'm running out of peak download allowance from my ISP and need a project that doesn't consume heaps (like BOINC Rosetta does).

Oh, they're real small. I promise. You can do tons of pairs. Just put all of your machines on NPLB and you'll have no problem at all.

Seriously, I have no idea. Max or David, can you help answer this question?

Quote:

Originally Posted by IronBits

That's funny about the servers getting sucked dry.

Let me know if it ever happens again so I can fix it for you.

Although it was a bit of a hassle, you're right, it is rather funny in retrospect.

What was it that Ross Perot said about jobs going south if NAFTA paased when he ran for president back in 1992: "That giant sucking sound that you'll hear is the stampede of jobs to Mexico if NAFTA passes". That's exactly what happened here. We got our pairs sucked south and we had to go get them back.

Alas, everything worked out.

Thanks for helping clear things back out again.

Gary

Mini-Geek · 2009-07-26, 10:50

Quote:

Originally Posted by vaughan

Can anyone help me with this question please?

How many kb are the NPLB tasks?
What would be the expected d/l kb per day for say an Intel Q6600?

I ask as I'm running out of peak download allowance from my ISP and need a project that doesn't consume heaps (like BOINC Rosetta does).

I don't really know, but I do know that the work units are tiny (e.g. drive 10's are 11 bytes: "kkkk nnnnnn"), so the biggest portion of the bandwidth is overhead in the communication. Is manual testing an option for this CPU? That would use up no bandwidth while it's running, and only a few dozen to hundred KB down/up every few days.

mdettweiler · 2009-07-26, 12:07

Quote:

Originally Posted by Mini-Geek

I don't really know, but I do know that the work units are tiny (e.g. drive 10's are 11 bytes: "kkkk nnnnnn"), so the biggest portion of the bandwidth is overhead in the communication. Is manual testing an option for this CPU? That would use up no bandwidth while it's running, and only a few dozen to hundred KB down/up every few days.

Yes, as Mini-Geek said. They're very small, but if you're really crunched for bandwidth, manual testing could help since it wouldn't require much more total bandwidth than you'd get by normally browsing the forum.

gd_barnes · 2009-07-27, 14:03

I will not be taking a business trip for at least 2 weeks now. Therefore I'm going to leave my G4000 and G8000 intact until about 2-3 days before I leave again for about a week. I'll be coordinating with David before that time to move them over to his servers. We can utilize port 3000 that has already been set up and will need to create a new one for the first 6k drive, perhaps port IB6000.

In the mean time, if there are any problems on my ports namely related to a power outage or my IP address changing, they should be resolved within a few hours instead of days or a week or more like last time. So if you want a 3-day testing window for returning pairs, you can have more confidence that they won't go down for an extended period.

Test servers and personal servers mostly used by Max but that he could also set up for others if needed will remain on my machine.

Gary

AMDave · 2009-07-28, 08:57

Thanks for the heads-up.
Implementation from the stats service should be auto-magic on appearance of the data from the ports on David's server.
I'll watch when it happens - weekend would be good ;) otherwise I might be at work.

2009-07-26, 00:48	#1125
mdettweiler A Sunny Moo Aug 2007 USA (GMT-5) 6249₁₀ Posts	I am pleased to announce that I've cleared up the mess on the GB servers, and have sent an email to David requesting that he do the same on the IB servers. As various users discovered earlier, it seems that some unknown person has sucked our servers dry of available work. No big deal; as soon as David's finished clearing things up on his end, we should be all back to normal. Let's keep those primes coming! Last fiddled with by mdettweiler on 2009-07-26 at 03:15

2009-07-26, 05:17	#1126
gd_barnes May 2007 Kansas; USA 24254₈ Posts	OK, everyone. The servers appear to be back to normal. We think that a new person accidentally cached a few too many pairs and dried us out temporarily. Everything looks good now. Thanks to David and Max for working diligently to get the problem resolved. Sorry about the problems. Gary Last fiddled with by gd_barnes on 2009-07-26 at 05:49

2009-07-26, 07:24	#1128
IronBits I ♥ BOINC! Oct 2002 Glendale, AZ. (USA) 3×7×53 Posts	That's funny about the servers getting sucked dry. Let me know if it ever happens again so I can fix it for you. Last fiddled with by IronBits on 2009-07-26 at 07:28

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
PRPnet servers for NPLB	mdettweiler	No Prime Left Behind	228	2018-12-26 04:50
Servers for NPLB	gd_barnes	No Prime Left Behind	0	2009-08-10 19:21
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
NPLB LLRnet server #1 - dried	em99010pepe	No Prime Left Behind	19	2008-03-26 06:19

2009-07-25, 20:49	#1124
mdettweiler A Sunny Moo Aug 2007 USA (GMT-5) 3·2,083 Posts	Here's the IP address that I was seeing on the G8000 console output. I can't be sure if this is the same one that pulled down all those k/n pairs since LLRnet doesn't correlate the two in its output, but here it is anyway: 97.222.133.84 (84.sub-97-222-133.myvzw.com) And its estimated location is: Country: UNITED STATES (US) City: Bedminster, NJ Anybody know of a guy from around the New Jersey area (or within the nearby states) who might even be remotely interested in prime searching or NPLB?

2009-07-26, 06:24	#1127
vaughan Jan 2005 Sydney, Australia 14F₁₆ Posts	Can anyone help me with this question please? How many kb are the NPLB tasks? What would be the expected d/l kb per day for say an Intel Q6600? I ask as I'm running out of peak download allowance from my ISP and need a project that doesn't consume heaps (like BOINC Rosetta does).

2009-07-27, 14:03	#1132
gd_barnes May 2007 Kansas; USA 2²×19×137 Posts	I will not be taking a business trip for at least 2 weeks now. Therefore I'm going to leave my G4000 and G8000 intact until about 2-3 days before I leave again for about a week. I'll be coordinating with David before that time to move them over to his servers. We can utilize port 3000 that has already been set up and will need to create a new one for the first 6k drive, perhaps port IB6000. In the mean time, if there are any problems on my ports namely related to a power outage or my IP address changing, they should be resolved within a few hours instead of days or a week or more like last time. So if you want a 3-day testing window for returning pairs, you can have more confidence that they won't go down for an extended period. Test servers and personal servers mostly used by Max but that he could also set up for others if needed will remain on my machine. Gary

2009-07-28, 08:57	#1133
AMDave Jan 2006 deep in a while-loop 1010010010₂ Posts	Thanks for the heads-up. Implementation from the stats service should be auto-magic on appearance of the data from the ports on David's server. I'll watch when it happens - weekend would be good ;) otherwise I might be at work.