Index  | Recent Threads  | Unanswered Threads  | Who's Active  | Guidelines  | Search
 

Quick Go »
No member browsing this thread
Thread Status: Active
Total posts in this thread: 14
Posts: 14   Pages: 2   [ Previous Page | 1 2 ]
[ Jump to Last Post ]
Post new Thread
Author
Previous Thread This topic has been viewed 4541 times and has 13 replies Next Thread
Sekerob
Ace Cruncher
Joined: Jul 24, 2005
Post Count: 20043
Status: Offline
Reply to this Post  Reply with Quote 
Re: HPF2 Longer runtimes and deadlines.

If Interested read up on Result Duration Correction Factor, sometimes abbreviated as RDCF but mostly DCF since the control parameter is called <duration_correction_factor>. It's the one the goes up very quick and goes down very slowly and acts as one of the gatekeepers to prevent clients from caching too much work. Except, here Alice (The Work Generator & Feeder & Scheduler Daemon) went off on a tangent and put wrong flop estimates into task headers. Like with a Jojo, what comes down, goes up again and overshoots if there is no one to catch it.

time for breakfast
----------------------------------------
WCG Global & Research > Make Proposal Help: Start Here!
Please help to make the Forums an enjoyable experience for All!
[Feb 29, 2008 7:16:17 AM]   Link   Report threatening or abusive post: please login first  Go to top 
Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline
Reply to this Post  Reply with Quote 
Re: HPF2 Longer runtimes and deadlines.

Hi all,
It might be because I am running on an intel based mac, but these new work units for HPF2 keep stalling for me around 98% (so far I have had three of five freeze and I let the three that failed each run stalled for over 4 days once they stopped moving). Any one have any ideas? I have done the standard suspend and resume to try to knock it out of the stall, but so far no luck. I haven't had any trouble with the pervious HPF2 work units so it just seems odd. Thanks!
[Feb 29, 2008 2:34:03 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Sekerob
Ace Cruncher
Joined: Jul 24, 2005
Post Count: 20043
Status: Offline
Reply to this Post  Reply with Quote 
Re: HPF2 Longer runtimes and deadlines.

Very precisely, suspend WCG in the Project Tab? Did you try fully exiting BOINc and restarting? Was it still accumulating CPU time in the Tasks Window and if not were the WCG sciences still consuming CPU time in the Process Manager view?

HPF2 has a build in time out if CPU time continues of 6x estimates flops (last update seen), so pretty much after 4 days they should have self aborted if consuming any.

The new lengthier HPF2 average was 9.52 hours yesterday. They run on Vista / Quad in range of 7 to 9 hours.
----------------------------------------
WCG Global & Research > Make Proposal Help: Start Here!
Please help to make the Forums an enjoyable experience for All!
[Feb 29, 2008 2:43:24 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline
Reply to this Post  Reply with Quote 
Re: HPF2 Longer runtimes and deadlines.

I did suspend WCG in the project tab and did restart BOINC. CPU time was not accumulating, but CPU time was still being consumed in the Process Manger view. That result is what threw me. For me, I don't really care about the points, but hate being someone who has to return an error. It could be a just a glitch in my computer. I am going to try removing BOINC and reinstalling the program to see if perhaps I had a bad install or something got corrupted. Just thought I would post the problem to see if it was unique to my computer. Thanks!
[Feb 29, 2008 3:02:45 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Posts: 14   Pages: 2   [ Previous Page | 1 2 ]
[ Jump to Last Post ]
Post new Thread