mersenneforum.org  

Go Back   mersenneforum.org > Great Internet Mersenne Prime Search > PrimeNet

Reply
 
Thread Tools
Old 2019-08-11, 17:21   #12
kriesel
 
kriesel's Avatar
 
"TF79LL86GIMPS96gpu17"
Mar 2017
US midwest

24×239 Posts
Default

patgie's 85575233 completed, leaving 28 that will expire in a day.
Unfortunately it was not well P-1 factored at the time. It has had additional sufficient P-1 factoring so the LL DC someday will not be wasted.
An unfortunate side effect of all those consecutive patgie reserve and expire cycles is it now looks as if patgie is assigned the DC of the LL first test completed by patgie.
kriesel is offline   Reply With Quote
Old 2019-08-14, 12:48   #13
Chuck
 
Chuck's Avatar
 
May 2011
Orange Park, FL

2×3×139 Posts
Default

He got all the same assignments back today. Why is this allowed to continue?
Chuck is offline   Reply With Quote
Old 2019-08-14, 17:38   #14
kriesel
 
kriesel's Avatar
 
"TF79LL86GIMPS96gpu17"
Mar 2017
US midwest

24·239 Posts
Default

Quote:
Originally Posted by Chuck View Post
He got all the same assignments back today. Why is this allowed to continue?
I don't know what patgie's renewal process is or how it is so effective. But patgie's down to 52 now from 58 earlier.
Patgie completed 4 in 5 days, and 2 were issued to others, who may or may not be faster than leaving them up to patgie.

There are at least two people well positioned to say what is going on; Madpoo after looking at server logs, and presumably patgie.

User WR is another somewhat interesting case in the first tests remaining before 86M milestone.

Code:
As of Aug 9, patgie had 58 exponents assigned to 86M first test milestone:
exponent    worktype stage % days left ETA       Assigned      Last update     Next Update      Est. Complete   UserID     status 2019-08-14
85532177    LL        3    15    2019-08-05    2019-08-05    2019-08-06    2019-08-24    patgie    done 8/12/19
85533977    LL        3    17    2019-08-05    2019-08-05    2019-08-06    2019-08-26    patgie    done 8/13/19
85550033    LL        3     7    2019-08-05    2019-08-05    2019-08-06    2019-08-16    patgie    assigned to Keith Schwab 8/14
85570697    LL        3    38    2019-08-05    2019-08-05    2019-08-06    2019-09-16    patgie    assigned to EthanD 8/14
85575233    LL        3     6    2019-08-05    2019-08-05    2019-08-06    2019-08-15    patgie    done 8/10/19
85583297    LL        3    15    2019-08-05    2019-08-05    2019-08-06    2019-08-24    patgie    pending reassigned 8/14
85585673    LL        3    34    2019-08-05    2019-08-05    2019-08-06    2019-09-12    patgie    done 8/10/19
85589681    LL        3     6    2019-08-05    2019-08-05    2019-08-06    2019-08-15    patgie    pending reassigned 8/14
85607897    LL        3    15    2019-08-05    2019-08-05    2019-08-06    2019-08-24    patgie    pending reassigned 8/14
85622177    LL        3    24    2019-08-05    2019-08-05    2019-08-06    2019-09-02    patgie    pending reassigned 8/14
85628513    LL        3     7    2019-08-05    2019-08-05    2019-08-06    2019-08-16    patgie    pending reassigned 8/14
85629737    LL        3    34    2019-08-05    2019-08-05    2019-08-06    2019-09-12    patgie    pending reassigned 8/14
85669601    LL        3     6    2019-08-05    2019-08-05    2019-08-06    2019-08-15    patgie    pending reassigned 8/14
85670393    LL        3     7    2019-08-05    2019-08-05    2019-08-06    2019-08-16    patgie    pending reassigned 8/14
85673201    LL        3    15    2019-08-05    2019-08-05    2019-08-06    2019-08-24    patgie    pending reassigned 8/14
85674521    LL        3    24    2019-08-05    2019-08-05    2019-08-06    2019-09-02    patgie    pending reassigned 8/14
85674881    LL        3    34    2019-08-05    2019-08-05    2019-08-06    2019-09-12    patgie    pending reassigned 8/14
85677377    LL        3     6    2019-08-05    2019-08-05    2019-08-06    2019-08-15    patgie    pending reassigned 8/14
85684721    LL        3    15    2019-08-05    2019-08-05    2019-08-06    2019-08-24    patgie    pending reassigned 8/14
85686521    LL        3    24    2019-08-05    2019-08-05    2019-08-06    2019-09-02    patgie    pending reassigned 8/14
85687817    LL        3    34    2019-08-05    2019-08-05    2019-08-06    2019-09-12    patgie    pending reassigned 8/14
85692521    LL        4     3    2019-08-06    2019-08-06    2019-08-07    2019-08-12    patgie    pending reassigned 8/14
85697321    LL        3     6    2019-08-05    2019-08-05    2019-08-06    2019-08-15    patgie    pending reassigned 8/14
85699433    LL        3    15    2019-08-05    2019-08-05    2019-08-06    2019-08-24    patgie    pending reassigned 8/14
85708793    LL        3    24    2019-08-05    2019-08-05    2019-08-06    2019-09-02    patgie    pending reassigned 8/14
85722473    LL        3    34    2019-08-05    2019-08-05    2019-08-06    2019-09-12    patgie    pending reassigned 8/14
85728641    LL        3     6    2019-08-05    2019-08-05    2019-08-06    2019-08-15    patgie    pending reassigned 8/14
85732553    LL        3    15    2019-08-05    2019-08-05    2019-08-06    2019-08-24    patgie    pending reassigned 8/14
85732601    LL        3    24    2019-08-05    2019-08-05    2019-08-06    2019-09-02    patgie    pending reassigned 8/14
85733201    LL        4     3    2019-08-06    2019-08-06    2019-08-07    2019-08-12    patgie    pending reassigned 8/14
85742897    LL        3    34    2019-08-05    2019-08-05    2019-08-06    2019-09-12    patgie    pending reassigned 8/14
85745921    LL        3     6    2019-08-05    2019-08-05    2019-08-06    2019-08-15    patgie    pending reassigned 8/14
85749953    LL        9     4    2019-07-29    2019-07-29    2019-07-30    2019-08-13    patgie    pending reassigned 7/29
85750337    LL        9     8    2019-07-29    2019-07-29    2019-07-30    2019-08-17    patgie    pending reassigned 7/29
85752977    LL        9    12    2019-07-29    2019-07-29    2019-07-30    2019-08-21    patgie    pending reassigned 7/29
85753001    LL        9    16    2019-07-29    2019-07-29    2019-07-30    2019-08-25    patgie    pending reassigned 7/29
85761833    LL        9    20    2019-07-29    2019-07-29    2019-07-30    2019-08-29    patgie    pending reassigned 7/29
85765793    LL    LL    79    -7    2019-07-29    2019-07-29    2019-07-30    2019-08-02    patgie    pending reassigned 7/29
85790153    LL        9     5    2019-07-29    2019-07-29    2019-07-30    2019-08-14    patgie    pending reassigned 7/29
85797617    LL        9     9    2019-07-29    2019-07-29    2019-07-30    2019-08-18    patgie    pending reassigned 7/29
85808081    LL        9    13    2019-07-29    2019-07-29    2019-07-30    2019-08-22    patgie    pending reassigned 7/29
85810601    LL        9    17    2019-07-29    2019-07-29    2019-07-30    2019-08-26    patgie    pending reassigned 7/29
85826057    LL        9    21    2019-07-29    2019-07-29    2019-07-30    2019-08-30    patgie    pending reassigned 7/29
85842857    LL    LL    79    -7    2019-07-29    2019-07-29    2019-07-30    2019-08-02    patgie    pending reassigned 7/29
85845377    LL        9    -3    2019-07-29    2019-07-29    2019-07-30    2019-08-06    patgie    pending reassigned 7/29
85855361    LL        9     1    2019-07-29    2019-07-29    2019-07-30    2019-08-10    patgie    pending reassigned 7/29
85893377    LL        9     5    2019-07-29    2019-07-29    2019-07-30    2019-08-14    patgie    pending reassigned 7/29
85893593    LL        9     9    2019-07-29    2019-07-29    2019-07-30    2019-08-18    patgie    pending reassigned 7/29
85898873    LL        9    13    2019-07-29    2019-07-29    2019-07-30    2019-08-22    patgie    pending reassigned 7/29
85910177    LL        9    17    2019-07-29    2019-07-29    2019-07-30    2019-08-26    patgie    pending reassigned 7/29
85910897    LL        9    21    2019-07-29    2019-07-29    2019-07-30    2019-08-30    patgie    pending reassigned 7/29
85919081    LL    LL    79    -7    2019-07-29    2019-07-29    2019-07-30    2019-08-02    patgie    pending reassigned 7/29
85938521    LL        9     5    2019-07-29    2019-07-29    2019-07-30    2019-08-14    patgie    pending reassigned 7/29
85943201    LL        9     9    2019-07-29    2019-07-29    2019-07-30    2019-08-18    patgie    pending reassigned 7/29
85967393    LL        9    -7    2019-07-29    2019-07-29    2019-07-30    2019-08-02    patgie    pending reassigned 7/29
85970561    LL        9    -3    2019-07-29    2019-07-29    2019-07-30    2019-08-06    patgie    pending reassigned 7/29
85972097    LL        9     1    2019-07-29    2019-07-29    2019-07-30    2019-08-10    patgie    pending reassigned 7/29
85976993    LL        9     1    2019-07-29    2019-07-29    2019-07-30    2019-08-10    patgie    pending reassigned 7/29

Last fiddled with by kriesel on 2019-08-14 at 17:41
kriesel is offline   Reply With Quote
Old 2019-08-14, 21:25   #15
Chuck
 
Chuck's Avatar
 
May 2011
Orange Park, FL

11010000102 Posts
Default

The ones he finished are assigned to him as double checks.

85575233
Chuck is offline   Reply With Quote
Old 2019-08-14, 22:15   #16
kriesel
 
kriesel's Avatar
 
"TF79LL86GIMPS96gpu17"
Mar 2017
US midwest

382410 Posts
Default

Quote:
Originally Posted by Chuck View Post
The ones he finished are assigned to him as double checks.

85575233
That should not happen. Note also that all the first time check assignments that expired in such a case got flipped to say double check now.

Last fiddled with by kriesel on 2019-08-14 at 22:17
kriesel is offline   Reply With Quote
Old 2019-08-18, 17:56   #17
hansl
 
hansl's Avatar
 
Apr 2019

3158 Posts
Default

Quote:
Originally Posted by kriesel View Post
That should not happen. Note also that all the first time check assignments that expired in such a case got flipped to say double check now.
Yeah this seems like a major problem/waste of CPU to me. Anything he double checks himself is going to have to be independently verified still, right?

Here's one that he's 75.6% done DC'ing his own first check:
https://www.mersenne.org/report_expo...5459097&full=1

Just checking exponent report for the upcoming 86M milestone:
https://www.mersenne.org/assignments...xp_hi=86000000

I clicked column to sort by userid and scrolled down to patgie, and I'm counting 76 total assignments, with 28 of those being double checks.

Out of those 28, 25 of them were initially checked by patgie!

Also it seems that these rules are not really enforced as described?
https://www.mersenne.org/thresholds/

For first time LL, Cat 2 and under specifies:
Quote:
Computer must have no expired assignments or bad or suspect results in the last 120 days.
Forget about Cat 2, all these <86M would be firmly in Cat 1 now, and should even be nearly all Cat 0 (Exponents below 85938522) at this point.

edit: Actually I guess since the check is per computer, maybe they could be swapping around which computer holds the assignments each time, to work around the check somehow? Doesn't seem that specific computer assigned is public info, but maybe someone with DB admin privileges could check if that's the case?

Last fiddled with by hansl on 2019-08-18 at 18:07
hansl is offline   Reply With Quote
Old 2019-08-21, 16:56   #18
kriesel
 
kriesel's Avatar
 
"TF79LL86GIMPS96gpu17"
Mar 2017
US midwest

73608 Posts
Default

Quote:
Originally Posted by hansl View Post
Actually I guess since the check is per computer, maybe they could be swapping around which computer holds the assignments each time, to work around the check somehow? Doesn't seem that specific computer assigned is public info, but maybe someone with DB admin privileges could check if that's the case?
It's "Computer: n/a" which is unusual. And might be how the no expirations on the computer restriction was being dodged. The reassignment code for first time checks and perhaps double checks has been recently modified by Prime95. It seems to have addressed the reassignment not following the assignment rules for first-time checks. Perhaps it also has for double-checks.

Quote:
Anything he double checks himself is going to have to be independently verified still, right?
Yes.

Last fiddled with by kriesel on 2019-08-21 at 16:58
kriesel is offline   Reply With Quote
Old 2019-08-21, 18:19   #19
Madpoo
Serpentine Vermin Jar
 
Madpoo's Avatar
 
Jul 2014

3×1,087 Posts
Default

Based on the theory that the user just happens to check in at roughly the same time, right after the assignments expire, I manually ran the daily expiration job at a different time last night, and again just now.

The patgie exponents that had "0 days" got expired, but here it is about 12 hours later and there were still quite a few of them available (17 I think). Some did get picked up by others.

I just ran it again ahead of schedule by a few hours and that expired the assignments to WR that were due to expire anyway when the job would normally do it's nightly run in about 5 hours. So now there are something like 41 exponents available in the 85M - 86M range.

Assuming those computers don't check in soon and re-grab those assignments (because they've already done some minimal amount of work on them), they should be available to others. So come and get 'em while you can.

If that's really all it is, then maybe we just need to alter the exact time of day when that job runs, in case this person deliberately set things up to check in at a specific time of day right after exponents expire.

But this also shows that even when they do expire, unless there are enough machines out there that fit the requirements for cat 0 work, they'll remain available for a good long period of time over the next day or two.
Madpoo is offline   Reply With Quote
Old 2019-08-21, 19:10   #20
kriesel
 
kriesel's Avatar
 
"TF79LL86GIMPS96gpu17"
Mar 2017
US midwest

24×239 Posts
Default

Quote:
Originally Posted by Madpoo View Post
Based on the theory that the user just happens to check in at roughly the same time, right after the assignments expire, I manually ran the daily expiration job at a different time last night, and again just now.

The patgie exponents that had "0 days" got expired, but here it is about 12 hours later and there were still quite a few of them available (17 I think). Some did get picked up by others.

I just ran it again ahead of schedule by a few hours and that expired the assignments to WR that were due to expire anyway when the job would normally do it's nightly run in about 5 hours. So now there are something like 41 exponents available in the 85M - 86M range.

Assuming those computers don't check in soon and re-grab those assignments (because they've already done some minimal amount of work on them), they should be available to others. So come and get 'em while you can.

If that's really all it is, then maybe we just need to alter the exact time of day when that job runs, in case this person deliberately set things up to check in at a specific time of day right after exponents expire.

But this also shows that even when they do expire, unless there are enough machines out there that fit the requirements for cat 0 work, they'll remain available for a good long period of time over the next day or two.
It's unclear to me whether it's due to your timing change, or George's code change. https://www.mersenneforum.org/showpo...postcount=3177 In any event, thank you both for all you do.
kriesel is offline   Reply With Quote
Old 2019-08-22, 00:14   #21
Chuck
 
Chuck's Avatar
 
May 2011
Orange Park, FL

2·3·139 Posts
Default

I wonder if user s0837263 is patgie under another name? All the same assignments back and all computer names are "n/a".
Chuck is offline   Reply With Quote
Old 2019-08-23, 23:31   #22
Madpoo
Serpentine Vermin Jar
 
Madpoo's Avatar
 
Jul 2014

3×1,087 Posts
Default

Quote:
Originally Posted by Chuck View Post
I wonder if user s0837263 is patgie under another name? All the same assignments back and all computer names are "n/a".
Entirely possible... brand new user created on Aug 21 but the assignments are older than that. I can imagine that happening if the new user was created and then took over the assignments.

Plus, there were other things that point heavily towards this being the same person.

It's unusual, I'll grant you that. The user "patgie" was not a new user. Turned in 1137 results (mostly LL/PRP) since it's creation nearly 3 years ago.

I'm still leaning towards thinking this is some weird issue with the client/software and nothing more. The new username looks like one of the auto-created ones? "S" and then a random #.

Whatever the case... next time they expire, one of y'all needs to step up and grab 'em quick. If the user is reading this forum, then maybe they'll be motivated to figure out what their machine(s) are doing, and if not, then there's no harm in me saying ahead of time that I'll try to run the expiration earlier than usual to give others more time to grab any.
Madpoo is offline   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
Add repeated digits after a number until it is prime sweety439 sweety439 73 2019-08-02 10:18
Completed assignments still in GPU72 Assignments page kladner GPU to 72 3 2019-06-15 01:19
Assignments Completed by Another User linament PrimeNet 99 2015-08-26 21:40
x86 Microcode Execution Unit and repeated-address implied loads ewmayer Hardware 8 2014-10-07 08:17
Trivial bug: repeated PM notifications Christenson Forum Feedback 0 2011-03-21 03:49

All times are UTC. The time now is 06:33.

Mon Jun 1 06:33:27 UTC 2020 up 68 days, 4:06, 1 user, load averages: 1.98, 2.78, 2.17

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2020, Jelsoft Enterprises Ltd.

This forum has received and complied with 0 (zero) government requests for information.

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.
A copy of the license is included in the FAQ.