mersenneforum.org  

Go Back   mersenneforum.org > New To GIMPS? Start Here! > Information & Answers

Reply
 
Thread Tools
Old 2012-05-17, 17:56   #1
ClownRoyal
 
ClownRoyal's Avatar
 
May 2012
Idaho

11002 Posts
Default Prime95 Mysteriously Abandoned Work

Hello,

Please help me get back my work. My quad-core runs 4 workers and last night, 3 of the workers inexplicably dropped what they were doing and moved on to different jobs.

I have three separate LL tests in various stages of completion - the largest (58mil+) was nearing completion. I was screwing around last night with mfaktc around the time that it happened, but I can't imagine what went wrong - it uses a separate directory/worktodo/results/etc.

I can still see the bu and bu2 files in my prime95 directory - actually three files each, one without extension. Browsing the forums, I found a similar post where a guy was able to modify the worktodo.txt but I don't know the exact syntax that should be entered for these exponents.

Exponents are listed below. Please help save my work if you can.

53109619
55421753
58056553

Thanks,

-cr


"Search on!"
ClownRoyal is offline   Reply With Quote
Old 2012-05-17, 19:22   #2
Dubslow
Basketry That Evening!
 
Dubslow's Avatar
 
"Bunslow the Bold"
Jun 2011
40<A<43 -89<O<-88

3·29·83 Posts
Default

Fortunately, all three exponents are still assigned to you. I have absolutely no idea why Prime95 would do something like that.

First, you must retrieve your Assignment ID for each exponent. There are two possible methods:

1) There is a program to automatically retrieve these. Find it here. (If it seems like a bunch of hassle, try doing it yourself first; it isn't really that hard.)

2) If that doesn't work or is too much hassle, you can do it manually, described below. (The automatic program does exactly this method, except... well, automatically.)

If you log in to PrimeNet and then go to your Assignments and then in your web browser select "View the source" (right click on the page, or one of the menus, or Ctrl+U in Google Chrome).

Roughly 500+ lines down, you'll see something like
Code:
<tr class="oddrow"><td>Gravemind</td><td align="center">1</td><td sorttable_customkey="54197029" align="right">54197029&nbsp;<input type="checkbox" name="hidden" value="R" style="width: 12; height: 12;"></td><td align="center">D</td><td align="right">LL,  82.40%</td><td>2012-05-07 18:46</td><td align="right">10</td><td>2012-05-17 02:56</td><td>2012-05-18 02:56</td><td>2012-05-19 13:06</td><td align="right">2</td></tr>
This is an example from my own account (this computer is named Gravemind). The difference is where I've written 'name="hidden"', you will see a long hex key; that is your AID. (Never post that on the forum.) The AID corresponds to the exponent which is written in the "sorttable_customkey" field, which is 54197029 in my example.

--------------------------------------------------------------------------------------

Having found the AID for each exponent you lost (doesn't matter which method), now you must shutdown Prime95 ("Test/Exit", don't just close the window) to modify the text files. Put the following line (based on my example in the manual method above) into worktodo.txt, one line for each exponent:
Code:
Test=AID,54197029,72,1
The first field is the AID, the second is the exponent; the third is the trial factoring depth (so it's 72 for two of those exponents, but 71 for another) and the last field is 0 for no P-1 complete or 1 for P-1 is complete. If you had already started the LL test for all three, then you can put ,72,1 for each of them and that won't screw anything else up. So, for example if my AID was DFCAB8000CFDE06B58203, it would look like
Code:
Test=DFCAB8000CFDE06B58203,54197029,72,1
You should also see the equivalent lines for whatever you are working on now; Prime95 will work on them in the order they are listed in the file, so if you want to resume these older ones first, put them ahead/above of the new exponents.
Dubslow is offline   Reply With Quote
Old 2012-05-17, 19:29   #3
ClownRoyal
 
ClownRoyal's Avatar
 
May 2012
Idaho

22·3 Posts
Default I think I answered my own question

In case anyone else ends up with the same problem, here is the fix that I found:

On Mersenne.org goto "results querries" then "exponent status" and check the box for "show exponents currently assigned to me" and fill in the appropriate exponent range.

Now you can see the results of trial factoring, etc. This is necessary for creating the correct syntax in your worktodo.txt. Keep this window open and open another window on mersenne.org.

Goto "my account" then "assignments." You will see a list of all exponents assigned to your various machines. View page source (google chrome shortcut CTRL+U) to find the hex key associated with each exponent.

A simple text search (CTRL+F) for your exponent will lead you to the code that describes it in the table. Find the entry in question and locate the large hexidecimal value called "checkbox name". Example:
"checkbox" name="7F528ECAB46C34B635BE38FA6FXXXXXX"

Now you have all the pieces of the puzzle and are ready to generate the command line for worktodo.txt which will set prime95 to work on your exponent.

Close prime95 before screwing around with the worktodo.txt file. Open worktodo.txt, select which worker you wish to assign to the job and add the entry under that worker's heading with this format:

test=HEXKEY,EXPONENT,NO_FACTORS_BELOW,1

HEXKEY - The large hexidecimal key you pulled from "my account", "assignments".
EXPONENT - The exponent of the mersenne candidate in question.
NO_FACTORS_BELOW - Value you pulled from the field of the same name on "results querries", "exponent status".
1 - This number tells something about how the exponent has been handled. I am unclear of the meaning and this can be taken as my best guess as to what to enter.

In the example below, there was already new work assigned to this worker, so I have inserted the job above the new assignment.

[Worker #2]
test=7F528ECAB46C34B635BE38FA6FXXXXXX,58570279,73,1
Test=85D59903BB17726B8851FD00D3XXXXXX,FFT2=2688K,51470249,72,1

Restart prime95 and it should pick up right where it left off, as long as you have all the necessary back up files from the previous work in the prime95 directory. It worked for me.

-cr

"Search On!"
ClownRoyal is offline   Reply With Quote
Old 2012-05-17, 19:33   #4
ClownRoyal
 
ClownRoyal's Avatar
 
May 2012
Idaho

22·3 Posts
Default slapping myself...

Quote:
Originally Posted by Dubslow View Post
Fortunately, all three exponents are still assigned to you. I have absolutely no idea why Prime95 would do something like that.
Gosh, thanks, Dubslow! You and I were both working hard typing away at the same time. I appreciate your reply and I hope I did not take up too much of your time.

-cr


"Search On!"
ClownRoyal is offline   Reply With Quote
Old 2012-05-17, 19:37   #5
Dubslow
Basketry That Evening!
 
Dubslow's Avatar
 
"Bunslow the Bold"
Jun 2011
40<A<43 -89<O<-88

3·29·83 Posts
Default

Quote:
Originally Posted by ClownRoyal View Post
1 - This number tells something about how the exponent has been handled. I am unclear of the meaning and this can be taken as my best guess as to what to enter.

As I said in my post above which you apparently missed while typing your second response, this indicates whether or not P-1 has been complete. In your case, all three have had P-1 complete. (If you had set 0, then Prime95 would have started a day or two of P-1 work before continuing the test.)

If I didn't know better I'd say you copied my post :razz: but at least you were able to work it out on your own. How did figure out where to find the AID in the page source? Did you see that elsewhere on the forum?


More cross posting!!!

Last fiddled with by Dubslow on 2012-05-17 at 19:37
Dubslow is offline   Reply With Quote
Old 2012-05-17, 19:48   #6
ClownRoyal
 
ClownRoyal's Avatar
 
May 2012
Idaho

22·3 Posts
Default CrossPosters Anonymous

Dubslow,

Yes, I saw another forum post where someone had used the "view source" method to get the AID but it did not state how to generate the rest of the test entry.

So I just kept poking around the forum and around mersenne.org until I figured out what to do with the rest of it.

Hey, out of curiosity; what's the worst that can happen if someone were to post their entire AID on the forum?

-cr

"Search On!"
ClownRoyal is offline   Reply With Quote
Old 2012-05-17, 20:22   #7
Dubslow
Basketry That Evening!
 
Dubslow's Avatar
 
"Bunslow the Bold"
Jun 2011
40<A<43 -89<O<-88

3×29×83 Posts
Default

It's sometimes possible to claim the assignment with the key; you could also unreserve the assignment and then whoever's working on it might run into problems submitting it. These are all highly unlikely, but hey.
Dubslow is offline   Reply With Quote
Old 2012-05-18, 05:06   #8
cheesehead
 
cheesehead's Avatar
 
"Richard B. Woods"
Aug 2002
Wisconsin USA

1E0C16 Posts
Default

Quote:
Originally Posted by ClownRoyal View Post
Hey, out of curiosity; what's the worst that can happen if someone were to post their entire AID on the forum?
Even if we can't think of much, just keep the AID confidential in case someone else does.
cheesehead is offline   Reply With Quote
Old 2012-05-18, 05:41   #9
LaurV
Romulan Interpreter
 
LaurV's Avatar
 
"name field"
Jun 2011
Thailand

100110111101012 Posts
Default

The worst can happen is someone else registers your exponent using your assignment key, you don't know and continue working, and find a prime. If I am the other guy I would say you poached me and I am 99% of work done (even if I didn't move a byte) and I may grab your $50k if I am enough clever.

When no prime is found, who cares about dry bones? Do you want my ~70GHzDays of credit? Be my guest, I won't cut your head for it. I made another in few days with a good GPU. But when the bone is juicy, all the dogs will start grrrrrrrrowlingggg...
LaurV is offline   Reply With Quote
Old 2012-05-18, 13:54   #10
Prime95
P90 years forever!
 
Prime95's Avatar
 
Aug 2002
Yeehaw, FL

41×193 Posts
Default

Are there any clues in results.txt or prime.log indicating why prime95 abandoned the exponents? If there is a bug, I need to fix it.
Prime95 is offline   Reply With Quote
Old 2012-05-19, 04:51   #11
ClownRoyal
 
ClownRoyal's Avatar
 
May 2012
Idaho

22×3 Posts
Default

Quote:
Originally Posted by Prime95 View Post
Are there any clues in results.txt or prime.log indicating why prime95 abandoned the exponents? If there is a bug, I need to fix it.
  • results.txt
    I do not see anything out of the ordinary in results.txt, nor any reference to the exponents in question.
  • prime.log
    I found nothing untoward in prime.log either. Just the usual "sending expected completion date...", "getting assignment from server..." etc. around the time in question. No reference to the abandoned exponents other than their original assignment and subsequent eta reports.

I wish I could offer more, but there just aren't any clues as far as my amateur eyes can tell.

@Prime95: Would you like to see the log file for yourself? I wouldn't put it past me to be missing something here. This system is still new to me.

-cr

"Search On!"
ClownRoyal is offline   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
Prime95 Work Preference tapion64 PrimeNet 14 2014-04-21 03:03
Stage 2 abandoned on shutdown smoffat Software 1 2008-11-21 20:56
release abandoned numbers? markg PrimeNet 3 2008-05-10 17:30
possibly abandoned repdigit prime project jasong jasong 8 2007-08-11 03:37
Question about abandoned tests birdman2584 Sierpinski/Riesel Base 5 3 2006-11-15 00:44

All times are UTC. The time now is 15:58.


Tue Jul 5 15:58:22 UTC 2022 up 82 days, 13:59, 1 user, load averages: 1.45, 1.47, 1.41

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2022, 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.

≠ ± ∓ ÷ × · − √ ‰ ⊗ ⊕ ⊖ ⊘ ⊙ ≤ ≥ ≦ ≧ ≨ ≩ ≺ ≻ ≼ ≽ ⊏ ⊐ ⊑ ⊒ ² ³ °
∠ ∟ ° ≅ ~ ‖ ⟂ ⫛
≡ ≜ ≈ ∝ ∞ ≪ ≫ ⌊⌋ ⌈⌉ ∘ ∏ ∐ ∑ ∧ ∨ ∩ ∪ ⨀ ⊕ ⊗ 𝖕 𝖖 𝖗 ⊲ ⊳
∅ ∖ ∁ ↦ ↣ ∩ ∪ ⊆ ⊂ ⊄ ⊊ ⊇ ⊃ ⊅ ⊋ ⊖ ∈ ∉ ∋ ∌ ℕ ℤ ℚ ℝ ℂ ℵ ℶ ℷ ℸ 𝓟
¬ ∨ ∧ ⊕ → ← ⇒ ⇐ ⇔ ∀ ∃ ∄ ∴ ∵ ⊤ ⊥ ⊢ ⊨ ⫤ ⊣ … ⋯ ⋮ ⋰ ⋱
∫ ∬ ∭ ∮ ∯ ∰ ∇ ∆ δ ∂ ℱ ℒ ℓ
𝛢𝛼 𝛣𝛽 𝛤𝛾 𝛥𝛿 𝛦𝜀𝜖 𝛧𝜁 𝛨𝜂 𝛩𝜃𝜗 𝛪𝜄 𝛫𝜅 𝛬𝜆 𝛭𝜇 𝛮𝜈 𝛯𝜉 𝛰𝜊 𝛱𝜋 𝛲𝜌 𝛴𝜎𝜍 𝛵𝜏 𝛶𝜐 𝛷𝜙𝜑 𝛸𝜒 𝛹𝜓 𝛺𝜔