| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: FightAIDS@Home
Thread: Have very slow WU # FAHV_x3VQ7_IN_LEDGFa_rig_0220728_0005 |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 234
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
A quick sample of the number of recent FAHV tasks per workunit on a PC: 37, 42, 99, 29, 41, 31, 51. All those WUs taking 0.5 to 0.7 hours.
|
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
On your advise revisited and there's indeed substantial variability in the packed jobs of fahv tasks. Did wcg finally implement the dynamic sizing to platform power or is this what in a beta thread is described as an estimator of the runtime duration, merely an object to dynamically get close to an mean time objective, no particular platform specific sizing?
A sample of runtime to jobs 6.39 - 39 4:94 - 45 5.94 - 33 4.52 - 38 5.32 - 35 5.85 - 32 Either way, it's a lot better than 100 hour jobs or what it was before 12-18 hours. Time and again find that if the charger can't keep up with the power consumption tasks fail with signal 11 or sigsegv. Then a short job crashing is lots less painful than a 12-18 hour. When seeing this, remotely launch a snooze with boinctasks, a function missing in the agent, so there's an hour postponed computing allowing the battery mostly to get back to max, which is when the charger can keep up, with screen off. |
||
|
|
littlepeaks
Veteran Cruncher USA Joined: Apr 28, 2007 Post Count: 748 Status: Offline Project Badges: 
|
I would like to know the chemical structural significance (if any) of these long running FAHV wus. In the graphics, it often shows the molecule separated into 3 or 4 submolecules, although I think this may just be an artifact of the graphic representation application.
Not receiving any more reruns. littlepeaks was curious |
||
|
|
KWSN - A Shrubbery
Master Cruncher Joined: Jan 8, 2006 Post Count: 1585 Status: Offline |
So the resends are still not done. Had one error out - maximum time exceeded today. That was copy 8, copy 9 has been sent.
----------------------------------------Another system is still running a resend. It's at 126 hours and closing in on 80%. I do hope the techs credit users for all this CPU time. It's high time the server just aborted any remaining long work units and do the calculations in house.  Distributed computing volunteer since September 27, 2000 |
||
|
|
Seoulpowergrid
Veteran Cruncher Joined: Apr 12, 2013 Post Count: 823 Status: Offline Project Badges: 
|
Had one error out - maximum time exceeded today. What is the maximum time? |
||
|
|
KWSN - A Shrubbery
Master Cruncher Joined: Jan 8, 2006 Post Count: 1585 Status: Offline |
The maximum time is calculated by flops. It varies wildly depending upon the speed of the machine.
----------------------------------------Distributed computing volunteer since September 27, 2000 |
||
|
|
Dennis-TW
Cruncher Joined: Apr 28, 2010 Post Count: 13 Status: Offline Project Badges: 
|
The maximum time is calculated by flops. It varies wildly depending upon the speed of the machine. That's strange. I have a WU with maximum time exceeded error after 38h, while half a dozen other WUs have been fine with 40+ some even with 50+ hours of runtime. All of them running on the same machine. |
||
|
|
TPCBF
Master Cruncher USA Joined: Jan 2, 2011 Post Count: 2173 Status: Offline Project Badges: 
|
For all I can tell, the issue with the overly long runtime WUs has been resolved, the returns both in numbers and in runtime on all of my systems seem to have returned to "pre-glitch" levels...
|
||
|
|
KWSN - A Shrubbery
Master Cruncher Joined: Jan 8, 2006 Post Count: 1585 Status: Offline |
For all I can tell, the issue with the overly long runtime WUs has been resolved, the returns both in numbers and in runtime on all of my systems seem to have returned to "pre-glitch" levels... Yes, they have. However, the bad batch is still kicking around in re-sends, hence the post. Distributed computing volunteer since September 27, 2000 |
||
|
|
Werinbert
Advanced Cruncher United States Joined: Jun 13, 2013 Post Count: 57 Status: Offline Project Badges:
|
200 hours wasted and abused! Not happy.
----------------------------------------My computer was happily crunching away on one of these long units. It takes about 250 hours to complete. Since it was a resend I got only three days to complete it, so it was sent out to other computers. One task happened to be sent to a fast computer completing it in two days. Well that task was returned and after a short while the scheduler removed the entire work unit from the system. My computer never had a chance. I crime was committed and the evidence was covered up.  [Edit 1 times, last edit by Werinbert at Oct 11, 2014 5:51:30 PM] |
||
|
|
|