OFFICIAL "SERVER PROBLEMS" THREAD - Page 125

Madpoo · 2017-12-17, 07:15

Quote:

Originally Posted by ET_

I am actually doing PRP-CF and PRP-CF-D.

Do you think I should stop until things are settled?

Luigi

Yeah, it was only an issue with older Prime95 versions that sent their PRP results to the server in a way that resulted in (sometimes) a missing list of known factors used during the test.

New versions (as George mentioned) will send the results fine.

tha · 2017-12-23, 20:26

I uploaded two results.txt files through the manual testing page. Among them there is the twin result from this exponent:

Code:

Manual testing	22268579	F-PM1	2017-12-23 19:32	0.0	Factor: 752894280511036848897308855040509839552636123889 / (P-1, B1=350000, B2=7000000, E=6)	0.8378
Manual testing	22268579	F-PM1	2017-12-23 19:32	0.0	Factor: 752894280511036848897308855040509839552636123889 / (P-1, B1=350000, B2=7000000, E=6)

unlike the other three factors that were uploaded in the same batch this twin result does not show up in the 'Current Progress | Recent Cleared' list, not when sorted on date/time stamp and not when sorted on exponent.
They do however show up correctly in the 'My Account | Results' page and therefore are correctly inserted into the database. Seems to be a report problem.

Prime95 · 2017-12-24, 00:04

Probably an artifact of the way composite factors are handled.

James Heinrich · 2017-12-24, 00:13

Composite factors will be broken into their prime component, but a side effect is that while the prime factors are recorded, the result log shows the original result text with the composite factor so the exponent detail page apparently shows the composite factor recorded twice. This is just a display artifact.

What I can't readily explain is why neither the prime nor composite factors, nor indeed that exponent at all, are shown on the recent cleared report.

potonono · 2018-01-10, 23:29

I have an automatic PRP assignment that seems to have expired in error. The CPU is set to get PRP-100M automatically for one worker, which it did.

https://www.mersenne.org/report_expo...2291569&full=1

2017-12-28 Assigned
2018-01-10 P1 results returned
2018-01-10 Expired

It is still in my local work-to-do for the remaining PRP work.

Prime95 · 2018-01-11, 02:22

Quote:

Originally Posted by potonono

I have an automatic PRP assignment that seems to have expired in error. The CPU is set to get PRP-100M automatically for one worker, which it did.

https://www.mersenne.org/report_expo...2291569&full=1

2017-12-28 Assigned
2018-01-10 P1 results returned
2018-01-10 Expired

It is still in my local work-to-do for the remaining PRP work.

Fixed. The P-1 result erroneously unassigned the exponent.

You should have the assignment back now.

S485122 · 2018-01-13, 08:29

It seems the first time checks category 0 threshold is stuck at 76737764 on the Assignment Rules page. This means there are only 10 category 0 LL assignments instead of 200.

Jacob

cuBerBruce · 2018-01-13, 14:21

Quote:

Originally Posted by S485122

It seems the first time checks category 0 threshold is stuck at 76737764 on the Assignment Rules page. This means there are only 10 category 0 LL assignments instead of 200.

Jacob

I count 11. You are probably not counting my "hidden" cat 0 (M76722991)

Prime95 · 2018-01-13, 17:25

Quote:

Originally Posted by S485122

It seems the first time checks category 0 threshold is stuck at 76737764 on the Assignment Rules page. This means there are only 10 category 0 LL assignments instead of 200.

Jacob

Fixed. The stored procedure was not recognizing exponents where a PRP result had been turned in.

GP2 · 2018-01-15, 12:56

I'm not sure why, but a user recently ran a PRP test (not a PRP cofactor test) on an already-factored small exponent: M3,859,447.

I think the server rejects LL residue results for already-factored exponents? If so, then probably the behavior should be consistent between LL results and PRP results.

PS,
The same user also ran a PRP test for a very similar exponent (M3,859,477) which doesn't have known factors, but did already have two matching LL tests. Probably this type of result should be accepted by the server, as it was, but it's an unverified result that doesn't need verification, similar to those few exponents in the 75M range (like M75,512,069) where we inadvertently ended up with two matching PRP tests and an unverified LL test. I presume the server already handles this situation and won't automatically assign a double-check on the superfluous test type?

tha · 2018-01-15, 16:16

On the page https://www.mersenne.org/report_recent_cleared/ the column 'Result' has been changed to include not just the result itself but also the method how it was obtained.

I have two suggestions.

A. Make the page wider, we now have only half the width of the screen in use and double lines because of the extra information not fitting in the cell.
B. Put the additional info on the method used in a separate cell so that automated use in spreadsheets and programs can use the results without having to undo the cell of the additional info.

2017-12-23, 20:26	#1366
tha Dec 2002 1456₈ Posts	I uploaded two results.txt files through the manual testing page. Among them there is the twin result from this exponent: Code: Manual testing 22268579 F-PM1 2017-12-23 19:32 0.0 Factor: 752894280511036848897308855040509839552636123889 / (P-1, B1=350000, B2=7000000, E=6) 0.8378 Manual testing 22268579 F-PM1 2017-12-23 19:32 0.0 Factor: 752894280511036848897308855040509839552636123889 / (P-1, B1=350000, B2=7000000, E=6) unlike the other three factors that were uploaded in the same batch this twin result does not show up in the 'Current Progress \| Recent Cleared' list, not when sorted on date/time stamp and not when sorted on exponent. They do however show up correctly in the 'My Account \| Results' page and therefore are correctly inserted into the database. Seems to be a report problem.

2018-01-10, 23:29	#1369
potonono Jun 2005 USA, IL 193 Posts	I have an automatic PRP assignment that seems to have expired in error. The CPU is set to get PRP-100M automatically for one worker, which it did. https://www.mersenne.org/report_expo...2291569&full=1 2017-12-28 Assigned 2018-01-10 P1 results returned 2018-01-10 Expired It is still in my local work-to-do for the remaining PRP work. Last fiddled with by potonono on 2018-01-11 at 00:12

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Official "Faits erronés dans de belles-lettres" thread	ewmayer	Lounge	39	2015-05-19 01:08
Official "all-Greek-to-me Fiction Literature and Cinema" Thread	ewmayer	Science & Technology	41	2014-04-16 11:54
Official "Lasciate ogne speranza" whinge-thread	cheesehead	Soap Box	56	2013-06-29 01:42
Official "Ernst is a deceiving bully and George is a meanie" thread	cheesehead	Soap Box	61	2013-06-11 04:30
Official "String copy Statement Considered Harmful" thread	Dubslow	Programming	19	2012-05-31 17:49

2017-12-24, 00:04	#1367
Prime95 P90 years forever! Aug 2002 Yeehaw, FL 7526₁₀ Posts	Probably an artifact of the way composite factors are handled.

2017-12-24, 00:13	#1368
James Heinrich "James Heinrich" May 2004 ex-Northern Ontario 3·5·227 Posts	Composite factors will be broken into their prime component, but a side effect is that while the prime factors are recorded, the result log shows the original result text with the composite factor so the exponent detail page apparently shows the composite factor recorded twice. This is just a display artifact. What I can't readily explain is why neither the prime nor composite factors, nor indeed that exponent at all, are shown on the recent cleared report.

2018-01-13, 08:29	#1371
S485122 Sep 2006 Brussels, Belgium 3226₈ Posts	It seems the first time checks category 0 threshold is stuck at 76737764 on the Assignment Rules page. This means there are only 10 category 0 LL assignments instead of 200. Jacob

2018-01-15, 12:56	#1374
GP2 Sep 2003 5·11·47 Posts	I'm not sure why, but a user recently ran a PRP test (not a PRP cofactor test) on an already-factored small exponent: M3,859,447. I think the server rejects LL residue results for already-factored exponents? If so, then probably the behavior should be consistent between LL results and PRP results. PS, The same user also ran a PRP test for a very similar exponent (M3,859,477) which doesn't have known factors, but did already have two matching LL tests. Probably this type of result should be accepted by the server, as it was, but it's an unverified result that doesn't need verification, similar to those few exponents in the 75M range (like M75,512,069) where we inadvertently ended up with two matching PRP tests and an unverified LL test. I presume the server already handles this situation and won't automatically assign a double-check on the superfluous test type?

2018-01-15, 16:16	#1375
tha Dec 2002 32E₁₆ Posts	On the page https://www.mersenne.org/report_recent_cleared/ the column 'Result' has been changed to include not just the result itself but also the method how it was obtained. I have two suggestions. A. Make the page wider, we now have only half the width of the screen in use and double lines because of the extra information not fitting in the cell. B. Put the additional info on the method used in a separate cell so that automated use in spreadsheets and programs can use the results without having to undo the cell of the additional info.