Results 1 to 7 of 7

Thread: Get Prime95 restarted on 17 or bust

  1. #1
    Registered User
    Join Date
    Jul 2012
    Location
    Brunswick, GA, USA
    Posts
    4

    Get Prime95 restarted on 17 or bust

    I'm running prime95 for Seventeen or Bust on three computers. I was running three instances on this computer. I had to make some changes with Windows last week. At the 17 or bust website, my production dropped, as if ones on this computer are not being counted. I went to Test/Primenet, and the user ID was blank. I tried to restart the program, but it started in GIMPS instead of 17 or bust.

    I lost my WorkToDo.txt file, but I have the data files. Is there a way I can get it to pick up where it left off in 17 or bust?

  2. #2
    Probably your configuration file (prime.txt) also was lost, so why user id was empty and server address set to default GIMPS. You could restore this from client archive (just overwrite existing one and set your username).

    You may TRY to continue already completed work if you have checkpoint files, named something like Z12345_678912345 (the first letter may be different, I don't remember which one exactly). Naming of them is very simple - first number is "k" and second is "n", so this file would mean workunit "12345*2^678912345+1".

    Find out which checkpoints you have. Then, go to http://seventeenorbust.com/account/testsPending.mhtml and check that these workunits are still assigned to you.

    To recover, get a copy of worktodo.txt from another computer. If you don't have one nearby, just let client to connect to SoB server and let it crunch for a few hours until it completes phase 1. Stop client before making any changes.

    Open worktodo.txt in simple text editor like notepad. You shall see some "PRP=...." lines. Their structure is very simple, it's easy to find which number corresponds to "12345" and "678912345" parts from example above. Duplicate, delete, move these lines as you wish, substituting "k" and "n" from names of checkpoint files you have. You may have any number of PRP=... lines, client will process them sequentially. When you start the client, it shall resume partial work from checkpoint.

  3. #3
    Registered User
    Join Date
    Jul 2012
    Location
    Brunswick, GA, USA
    Posts
    4
    Sorry it took me a while to respond, I thought I had auto notifications on, but I didn't.

    I found the three it was working on and they are still assigned to me (they all show 0 progress though, on that page - which isn't right - one was 98% done). A few days ago I got it to restart on new SoB tests. Here is an example of the worktodo file:

    [Worker #1]
    PRP=0000000000000000000000000011D160,24737,2,24611863,1
    PRP=0000000000000000000000000011D913,55459,2,24693814,1,56,2.1

    So most of it is clear to me - the first one is the one it is working on and the second one is in the queue. I don't know what the 56,2.1 is, but all of the ones in the queue have that.

    But what about the PRP number? Do I need that? What is it - the hexadecimal value of the ID number, or something?
    Last edited by JudMcCranie; 05-04-2014 at 01:50 AM.

  4. #4
    Quote Originally Posted by JudMcCranie View Post
    So most of it is clear to me - the first one is the one it is working on and the second one is in the queue. I don't know what the 56,2.1 is, but all of the ones in the queue have that.
    They're all in the queue, but the first line has already completed Phase 1 (factoring) test, while second (which wasn't started yet) did not. These numbers are parameters for factoring test. You should not use them for restored work.

    But what about the PRP number? Do I need that? What is it - the hexadecimal value of the ID number, or something?
    Yes, this is the hexadecimal value of the ID from pending tests page. I think that all of ID, k, n must match, so get correct value from page and convert it to hexadecimal.

  5. #5
    Registered User
    Join Date
    Jul 2012
    Location
    Brunswick, GA, USA
    Posts
    4
    OK, I think I can get it done. (In a couple of days.) Thanks.

  6. #6
    Well, good luck. And one last point which I've remembered. You may have two kinds of checkpoint files. Important ones from PRP phase (time-consuming). They're named as P###_###### (note the first letter). For them you must use first type of line, like this (from your example):

    PRP=0000000000000000000000000011D160,24737,2,24611863,1

    You also may have checkpoints from factoring phase. They're named M###_#####. Actually, you'll not get much from recovering them - factoring takes only about 6 hours on average PC, so it could be simpler just to drop them and cancel assignment on pending tests page. If you still want to recover them, use second type of line, with factoring parameters:

    PRP=0000000000000000000000000011D913,55459,2,24693814,1,56,2.1

    Don't forget to close client before making changes, and backup whole directory first.

  7. #7
    Registered User
    Join Date
    Jul 2012
    Location
    Brunswick, GA, USA
    Posts
    4
    Well, looks like I got it done! One of the tests was 98% finished - needs less than 2.5 hours to finish. The other two weren't very far along.

    Thanks!

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •