• March 06, 2021, 06:11:07 AM
  • Welcome, Guest
Please login or register.

Login with username, password and session length
Advanced search  


We have a Twiiter Account for the Team

Pages: 1 [2]   Go Down

Author Topic: Possible work from LHC  (Read 56394 times)


  • Information Member
  • Board Lord
  • *****
  • Aussie Karma: +932/-0
  • Offline Offline
  • Posts: 2746
Re: Possible work from LHC
« Reply #15 on: February 01, 2018, 06:36:15 PM »

Yes EVENTUALLY they seem to arrive (WU's)
But who can sit around with idle cores waiting! :D


  • Forum Moderator
  • Board Lord
  • *****
  • Aussie Karma: +690/-0
  • Offline Offline
  • Posts: 2893
Re: Possible work from LHC
« Reply #16 on: February 01, 2018, 11:37:52 PM »

Yes, put LHC Sixtrack on Skylake box and noticed it ran out of work once. So attached to WCG with a resource share preference set to zero and it downloads some OET tasks if LHC runs dry. But it only downloads enough OET tasks to fill any cores that are idle. With cache set to 4 days when Sixtrack tasks are available it gets some then. Finishes Sixtrack tasks much quicker than other box though and/or has downloaded more "shorties" so cache is often lowish, currently only 28.

Dual socket box downloaded 480 Sixtrack tasks and has not looked like running out of Sixtrack work since. It tops up now and again and whenever I've looked I've never seen less than 400 Sixtrack tasks in cache (468 right now). Currently have only WCG Help Stop TB also selected on that box with a high resource share and actually just got 4 tasks, the first Help Stop TB tasks it's managed to snag in months.

Edit: Hmm, Skylake uploaded its last Linux Sixtrack task at 4.20 am and hasn't received any since then. At least it's not idle as WCG stepped in automatically to fill the void. Strangely duallie still has 461 Windows tasks in cache. Will try swapping Skylake back to native Windows and see if work availability increases. Haven't had any prolonged upload sticking so at least that's working.
« Last Edit: February 02, 2018, 12:42:42 PM by kashi »
Pages: 1 [2]   Go Up

Page created in 0.056 seconds with 21 queries.