1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

Bigadv adventures

Discussion in 'Team OcUK Distributed Computing Projects' started by Senture, 14 Apr 2010.

  1. Senture

    Wise Guy

    Joined: 1 May 2007

    Posts: 2,355

    Location: London

    Got the use of an 8 core 3ghz Xeon machine with 8GB of ram for a short while, but no hard disk. I managed to install Ubuntu 10.04 onto my 4GB Corsair Flash Voyager and had to strip out some unneeded applications to get FAH stable (lost a couple of bigadv units early on :( never mind). Completed the first unit about an hour and a half ago, WU Results file compiled but it hasn't started the upload process yet. It's been molesting my Flash Voyager the whole time and shows no sign of stopping.

    Don't have any hard drives I can use in the machine, is there any way I could PXE boot the entire OS from another machine?. Also, should I consider terminating the FAH process or am I likely to brick the WU?
     
  2. Biffa

    Man of Honour

    Joined: 30 Oct 2002

    Posts: 15,628

    Location: Surrey

    Whats the frame time like on that?

    You should be able to ctrl^c the process without fubar'ing the WU
     
  3. Senture

    Wise Guy

    Joined: 1 May 2007

    Posts: 2,355

    Location: London

    They're only X5450's so not as great as the 55xx series. Just under 33 mins TPF which isn't great but sure better that the 4K PPD the pair of 5160's in the other machine can produce. WU is uploading now, took 2 and a half hours to begin the upload, that's going to put a dent in the final output.
     
  4. Biffa

    Man of Honour

    Joined: 30 Oct 2002

    Posts: 15,628

    Location: Surrey

    Ah so its the Xeon version of the Q9650? Sounds about right then for the TPF.

    2 and a half hours??? :eek: My first bigadv took 15 minutes to save and upload.

    Code:
    [09:54:38] Finished Work Unit:
    [09:54:38] - Reading up to 52713120 from "work/wudata_02.trr": Read 52713120
    [09:54:39] trr file hash check passed.
    [09:54:39] - Reading up to 42831368 from "work/wudata_02.xtc": Read 42831368
    [09:54:39] xtc file hash check passed.
    [09:54:39] edr file hash check passed.
    [09:54:39] logfile size: 209680
    [09:54:39] Leaving Run
    [09:54:40] - Writing 95919476 bytes of core data to disk...
    [09:54:44]   ... Done.
    [09:56:11] - Shutting down core
    [09:56:11] 
    [09:56:11] [email protected] Core Shutdown: FINISHED_UNIT
    [09:58:05] CoreStatus = 64 (100)
    [09:58:05] Unit 2 finished with 54 percent of time to deadline remaining.
    [09:58:05] Updated performance fraction: 0.539549
    [09:58:05] Sending work to server
    [09:58:05] Project: 2681 (Run 10, Clone 5, Gen 70)
    [09:58:05] + Attempting to send results [April 13 09:58:05 UTC]
    [09:58:05] - Reading file work/wuresults_02.dat from core
    [09:58:06]   (Read 95919476 bytes from disk)
    [09:58:06] Connecting to http://171.67.108.22:8080/
    [10:09:50] Posted data.
    [10:09:50] Initial: 0000; - Uploaded at ~133 kB/s
    [10:09:50] - Averaged speed for that direction ~133 kB/s
    [10:09:50] + Results successfully sent
    [10:09:50] Thank you for your contribution to [email protected]
    
    
     
  5. Senture

    Wise Guy

    Joined: 1 May 2007

    Posts: 2,355

    Location: London

    Sure is taking me a lot longer, but this is possibly due to running it off a pen drive.

    Code:
    [04:52:14] Completed 240000 out of 250000 steps  (96%)
    [05:25:13] Completed 242500 out of 250000 steps  (97%)
    [05:58:10] Completed 245000 out of 250000 steps  (98%)
    [06:31:09] Completed 247500 out of 250000 steps  (99%)
    [07:04:22] Completed 250000 out of 250000 steps  (100%)
    [07:04:29] DynamicWrapper: Finished Work Unit: sleep=10000
    [07:04:39] 
    [07:04:39] Finished Work Unit:
    [07:04:39] - Reading up to 52713120 from "work/wudata_01.trr": Read 52713120
    [07:04:39] trr file hash check passed.
    [07:04:39] - Reading up to 47064260 from "work/wudata_01.xtc": Read 47064260
    [07:04:40] xtc file hash check passed.
    [07:04:40] edr file hash check passed.
    [07:04:40] logfile size: 235126
    [07:04:40] Leaving Run
    [07:04:44] - Writing 100177422 bytes of core data to disk...
    [07:04:49]   ... Done.
    [09:23:45] - Shutting down core
    [09:23:45] 
    [09:23:45] [email protected] Core Shutdown: FINISHED_UNIT
    [09:24:13] CoreStatus = 64 (100)
    [09:24:13] Sending work to server
    [09:24:13] Project: 2681 (Run 14, Clone 17, Gen 47)
    
    
    [09:24:13] + Attempting to send results [April 14 09:24:13 UTC]
    [09:48:56] + Results successfully sent
    [09:48:56] Thank you for your contribution to [email protected]
    [09:48:56] + Starting local stats count at 1
    [10:08:10] - Preparing to get new work unit...
    [10:08:10] Cleaning up work directory
    [10:08:10] + Attempting to get work packet
    [10:08:10] Passkey found
    [10:08:10] - Connecting to assignment server
    [10:08:11] - Successful: assigned to (171.67.108.22).
    [10:08:11] + News From [email protected]: Welcome to [email protected]
    
     
  6. Marine Iguana

    Soldato

    Joined: 25 Sep 2008

    Posts: 6,769

    Location: Orsett, Essex

    Hmm i have only recently overtaken yourself and now this pfft got a whole swarm up my rear :p
     
  7. Senture

    Wise Guy

    Joined: 1 May 2007

    Posts: 2,355

    Location: London

    Haha :p. Not going to have access to the machine for long, just fancied devouring a few big WUs :D
     
  8. Senture

    Wise Guy

    Joined: 1 May 2007

    Posts: 2,355

    Location: London

    Oh bugger, looks like I destroyed Ubuntu removing packages because of insufficient space :(
     
  9. Senture

    Wise Guy

    Joined: 1 May 2007

    Posts: 2,355

    Location: London

    Ok I hate linux now. Backed up the entire [email protected] client onto my VM last night and reinstalled Ubuntu. Left the WU running after doing all updates and left it to complete the last 1%. Client threw a wobbly and deleted the WU, then the package removal process threw a wobbly about disk space and corrupted xorg and the man-db item :(

    Attempt #2 today, and fah bricked the WU again for no reason. And it's failing to get work :mad:
     
  10. Senture

    Wise Guy

    Joined: 1 May 2007

    Posts: 2,355

    Location: London

    Right I give up with bigadv folding, it's just not worth it. Completed my second unit after a week, no credit given. Posted about it on the folding forum, WU had no passkey (incorrect). Next WU bricks from insufficient space, 4th unit bricks for no reason, 5th bricks after a clean shutdown and 6th bricks randomly.

    System passes every single test I've thrown at it but for no reason, WU's keep bricking more commonly that Xbox 360's :o. Enough of this, Windows 7 installed and SMP client running as a service stable. If any one has any magical ideas to get bigadv WUs to not brick, I'm all ears. The extra 8k PPD would be nice, but if it takes me a week to complete one due to multiple failures and possible zero credit, to hell with it.
     
  11. Biffa

    Man of Honour

    Joined: 30 Oct 2002

    Posts: 15,628

    Location: Surrey

    Sounds familiar.
     
  12. rich99million

    Sgarrista

    Joined: 26 Dec 2002

    Posts: 9,348

    Location: Derbyshire

    Woah what just happened? My rearview had you down as just under a month behind me and it's jumped to just over a week and a half - bleedin new-fangled massive WUs :p :D

    Looks like a short excursion outside the top 30 is on the cards for me then - until I catch up with Milky^ in a month or so - need more powahhhh!!!
     
  13. Senture

    Wise Guy

    Joined: 1 May 2007

    Posts: 2,355

    Location: London

    :o didn't realize that was so close. Had another bash with Bigadv after some one pointed out on folding forums that the cause of this issue of mine may be the Ext4 file system, nuked the windows install and now running Ubuntu 9.10 with kernel 2.6.33.2 getting 32:50 TPF stable. Not tried to kill the process yet but no need so far.

    I'll be nice and issue with a parp rich, welcome back btw :D
     
  14. rich99million

    Sgarrista

    Joined: 26 Dec 2002

    Posts: 9,348

    Location: Derbyshire

    Glad you got it sorted - even if it does mean I have to move a cell in the wrong direction for a bit :p
     
  15. Marine Iguana

    Soldato

    Joined: 25 Sep 2008

    Posts: 6,769

    Location: Orsett, Essex

    Never knew how hard it is up here in the elite part of the team get some upstarts trying to take your hard earned places eh ? :p:D
     
  16. VeNT

    Capodecina

    Joined: 9 Jan 2003

    Posts: 20,703

    Location: Cornwall

    what kinda drive does it take?
     
  17. Senture

    Wise Guy

    Joined: 1 May 2007

    Posts: 2,355

    Location: London

    The 8 core machine I have? Can either put SAS or SATA drives in there. Only had a SATA drive laying about so that's in use :)
     
  18. VeNT

    Capodecina

    Joined: 9 Jan 2003

    Posts: 20,703

    Location: Cornwall

    wonder if there is anyone with a spair SATA in london they could loan you.
     
  19. Senture

    Wise Guy

    Joined: 1 May 2007

    Posts: 2,355

    Location: London

    Luckily I found an 80GB SATA drive kicking about so that's in use. Stuck Windows 7 on there for a while whilst I got over my anger at the client, reinstalled Ubuntu 9.10 and changed the file system from Ext4 to JFS as it had been noted on the folding forum that this resolves the issue I had, and it sure did. Write times down from 2.5 hours to 15 mins (still slow, but never mind) and after 4 power cuts, I've yet to have a WU brick :D
     
  20. Senture

    Wise Guy

    Joined: 1 May 2007

    Posts: 2,355

    Location: London

    Just a courtesy reminder to tidy up the cell before you relocate :). EOC think it will be 1.9 days but the next 57k WU completes in 22 hours :)