Linear algebra reservations, progress and results - Page 16

RichD · 2015-04-28, 22:20

as ...

Code:

prp71 factor: 35624182496410965748829117789675517088432536079668946359416981615199329
prp166 factor: 1123197585269685489106584886107619791236398381613505787213634888393640062353436618097006764445712431926236343004917746201147939369386304168491784104001862759169278369

70 hours solving a 9.2M matrix on Core-i5 using TD=124 & -t 4.

RichD · 2015-04-30, 22:14

I'll take GW_7_291 & GW_7_292 next.

YuL · 2015-05-01, 09:25

Quote:

Originally Posted by YuL

I can deal with GW_6_317 next.

Matrix successfully built @td=116. LA in progress:

Code:

linear algebra completed 1546945 of 15717685 dimensions (9.8%, ETA 108h50m)

VictordeHolland · 2015-05-01, 14:39

Due to an ever increasing powerbill, I've decided to shut down a 3770k and take a small break from post-processing. This also gives me some time to do rewiring of my Ethernet and clean up the cable mess. I expect to be back in a couple of weeks.

RichD · 2015-05-01, 14:43

Quote:

Originally Posted by RichD

I'll take GW_7_291 & GW_7_292 next.

Both have their matrices built and should complete with LA at the end of next week.

jyb · 2015-05-01, 17:58

Can somebody explain the following information on the 14e status page for GW_10_244 and GW_10_247? Little of it makes sense to me.

Code:

#	Name		Project	Type		Bits	Q range		Pushed	Unsent	Pending	Received	Relations	Est. Pending Rels
8	GW_10_244	GCW	SNFS(246)	31	20-100 M	100 %	0	21	4086		94611657	21163743
9	GW_10_247	GCW	SNFS(249)	31	20-200 M	100 %	2	1387	995		24365779	251077720

Specifically,

10,244:
- With 21 work units pending, how can there possibly be anywhere close to 21M pending relations?
- Why was the special-q range chosen with such a low upper limit? Is there any expectation that 120M or so relations will be enough?

10,247:
- What happened to all the work units? There were 11250 created, and many seem to have disappeared.
- Received has been stuck on 995 for days, though I know that many have been uploaded in that time.
- The pending relations estimate is completely out of whack with the number of pending work units (as with 10,244).

Does some server just need a good kick?

VictordeHolland · 2015-05-01, 19:44

Quote:

Originally Posted by jyb

Does some server just need a good kick?

It wouldn't be the first time the status update script took a little nap and needed a little love.
Normally I wouldn't thrust the estimated pending relations, but the gathered raw relations is fairly accurate (if the script works).
However it is not unusual to first push 100M special Q to see how many relations are collected. Based on that the range can be extended to achieve the desired amount.

jyb · 2015-05-04, 17:52

Quote:

Originally Posted by VictordeHolland

It wouldn't be the first time the status update script took a little nap and needed a little love.
Normally I wouldn't thrust the estimated pending relations, but the gathered raw relations is fairly accurate (if the script works).
However it is not unusual to first push 100M special Q to see how many relations are collected. Based on that the range can be extended to achieve the desired amount.

Okay, well AFAICT none of the currently sieving numbers has had its relation count updated since April 27th. And that's not just the status page; the directories for each number in the nfs_data directory haven't been updated since then either. It looks like *something* needs a little love.

frmky · 2015-05-04, 23:23

Updates in progress now...

fivemack · 2015-05-05, 07:19

Starting linear algebra on C170_146_126

fivemack · 2015-05-05, 09:05

Code:

Sun May  3 21:45:00 2015  prp75 factor: 207411584559018609890401114827572816371627045253302551111130346462621961989
Sun May  3 21:45:00 2015  prp158 factor: 18404143479135655578778200998433256754406015243619955924279549245401838098514089562032266716711141389520401463898803464243560647281911089396643278629387591653

~216 hours for 13.9M density-110 matrix on i7/2600 -t3. Log attached.

2015-05-01, 17:58	#171
jyb Aug 2005 Seattle, WA 2·883 Posts	Can somebody explain the following information on the 14e status page for GW_10_244 and GW_10_247? Little of it makes sense to me. Code: # Name Project Type Bits Q range Pushed Unsent Pending Received Relations Est. Pending Rels 8 GW_10_244 GCW SNFS(246) 31 20-100 M 100 % 0 21 4086 94611657 21163743 9 GW_10_247 GCW SNFS(249) 31 20-200 M 100 % 2 1387 995 24365779 251077720 Specifically, 10,244: - With 21 work units pending, how can there possibly be anywhere close to 21M pending relations? - Why was the special-q range chosen with such a low upper limit? Is there any expectation that 120M or so relations will be enough? 10,247: - What happened to all the work units? There were 11250 created, and many seem to have disappeared. - Received has been stuck on 995 for days, though I know that many have been uploaded in that time. - The pending relations estimate is completely out of whack with the number of pending work units (as with 10,244). Does some server just need a good kick?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
restarting nfs linear algebra	cubaq	YAFU	2	2017-04-02 11:35
Linear algebra at 600%	CRGreathouse	Msieve	8	2009-08-05 07:25
Linear algebra crashes	10metreh	Msieve	3	2009-02-02 08:34
Linear algebra proof	Damian	Math	8	2007-02-12 22:25
Linear algebra in MPQS	R1zZ1	Factoring	2	2007-02-02 06:45

2015-04-30, 22:14	#167
RichD Sep 2008 Kansas 2·3·5·113 Posts	I'll take GW_7_291 & GW_7_292 next.

2015-05-01, 14:39	#169
VictordeHolland "Victor de Hollander" Aug 2011 the Netherlands 2³·3·7² Posts	Due to an ever increasing powerbill, I've decided to shut down a 3770k and take a small break from post-processing. This also gives me some time to do rewiring of my Ethernet and clean up the cable mess. I expect to be back in a couple of weeks.

2015-05-04, 23:23	#174
frmky Jul 2003 So Cal 84B₁₆ Posts	Updates in progress now...

2015-05-05, 07:19	#175
fivemack (loop (#_fork)) Feb 2006 Cambridge, England 1100100011000₂ Posts	Starting linear algebra on C170_146_126