| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Support
Forum: BOINC Agent Support
Thread: I hope BOINC can fix these Bugs ASAP. |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 10
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
My computer needs over 10 hours CPU time(run almost 30 hours) to
complete a HCC/hpf/faah Work unit and Boinc makes a checkpoint every 60~120 minutes. It's not fast, but I think WCG still have many members get similar results. But if I turn off/restart my computer or snooze process or Boinc run a CPU benchmark(default setting) before a new checkpoint has been made, current Boinc will lose all the efforts made after previous checkpoint, 2 hours gone. All BOINCed members affected. I know my computer can choose DDDt to do, but my concern is not how long a work unit takes, is how long Boinc wastes. So I hope new Boinc can save all jobs before exit and can manually make a new checkpoint and make "leave applications in memory while suspended" as defoult setting, such improvement can save every minute we contributed, thanks. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
This behaviour is by design.
It's the best we can manage. Here's why: http://www.worldcommunitygrid.org/forums/wcg/viewthread?thread=11332 |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
This obscures me
----------------------------------------.... run a CPU benchmark(default setting) before a new checkpoint has been made, current Boinc will lose all the efforts made after previous checkpoint, 2 hours gone. as it is behavior of an old <= 5.4.13 BOINC version. The current 5.10 version, but also 5.8 should either through manual or scheduled benchmarking pre-empt the calculation and leave it in memory automatically and just resumes calculation without a second loss. That said, I've seen with 5.10.42 a strange case where a high priority job pre-empted a regular job. After the HP job finished, the regular job fell back several percentage points and lost about 15 minutes progress. Maybe Didactylos knows about this.... LIM was off as i checked that right after?
WCG
----------------------------------------Please help to make the Forums an enjoyable experience for All! [Edit 1 times, last edit by Sekerob at Feb 23, 2008 9:18:40 PM] |
||
|
|
retsof
Former Community Advisor USA Joined: Jul 31, 2005 Post Count: 6824 Status: Offline Project Badges: 
|
If you specify for BOINC to leave jobs in memory, wouldn't the other regular job continue where it left off? (If the computer is turned off, this wouldn't apply.)
----------------------------------------My computer needs over 10 hours CPU time(run almost 30 hours) to complete a HCC/hpf/faah Work unit and Boinc makes a checkpoint every 60~120 minutes. I also wonder whether his BOINC throttle is set to something other than 100%... It may be an issue for a laptop, but a desktop might as will run as much as possible. But if I turn off/restart my computer or snooze process or Boinc run a CPU benchmark(default setting) before a new checkpoint has been made, current Boinc will lose all the efforts made after previous Either try a different project with more checkpoints, or let the computer run. Checkpoints are then not a problem at all.checkpoint, 2 hours gone.
SUPPORT ADVISOR
Work+GPU i7 8700 12threads School i7 4770 8threads Default+GPU Ryzen 7 3700X 16threads Ryzen 7 3800X 16 threads Ryzen 9 3900X 24threads Home i7 3540M 4threads50% |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
It's no loss if you checked "Leave applications in memory while suspended" option.
This obscures me .... run a CPU benchmark(default setting) before a new checkpoint has been made, current Boinc will lose all the efforts made after previous checkpoint, 2 hours gone. as it is behavior of an old <= 5.4.13 BOINC version. The current 5.10 version, but also 5.8 should either through manual or scheduled benchmarking pre-empt the calculation and leave it in memory automatically and just resumes calculation without a second loss. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I also wonder whether his BOINC throttle is set to something other than 100%... It may be an issue for a laptop, but a desktop might as will run as much as possible. But if I turn off/restart my computer or snooze process or Boinc run a CPU benchmark(default setting) before a new checkpoint has been made, current Boinc will lose all the efforts made after previous Either try a different project with more checkpoints, or let the computer run. Checkpoints are then not a problem at all.checkpoint, 2 hours gone. Checkpoint is really important for people who running projects with limited cup time. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
thanks.
Can we suggest designer to fix it in next release? almost one million devices will save huge time. This behaviour is by design. It's the best we can manage. Here's why: http://www.worldcommunitygrid.org/forums/wcg/viewthread?thread=11332 |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
As explained in the FAQ, can't be done unless a severe price is paid in reduction in user experience.
----------------------------------------You don't tell which version of BOINC u use, but advise you upgrade to 5.10.42 or later version when WCG declares it 'recommended', by making it available through their download page.
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
retsof
Former Community Advisor USA Joined: Jul 31, 2005 Post Count: 6824 Status: Offline Project Badges:
|
How much memory is on that computer, what is the OS, how fast is the computer and what is the BOINC version?
----------------------------------------We're not sure why it is so slow between checkpoints unless it is a very old machine. A 512MHz Win98 machine here has about given up so it is off of the project. My AMD64 2.4GHz are taking a checkpoint about every minute on DDD-T. There's really not much loss, even if the computers were turned off at random.
SUPPORT ADVISOR
----------------------------------------Work+GPU i7 8700 12threads School i7 4770 8threads Default+GPU Ryzen 7 3700X 16threads Ryzen 7 3800X 16 threads Ryzen 9 3900X 24threads Home i7 3540M 4threads50% [Edit 1 times, last edit by retsof at Feb 24, 2008 1:10:12 PM] |
||
|
|
JmBoullier
Former Community Advisor Normandy - France Joined: Jan 26, 2007 Post Count: 3716 Status: Offline Project Badges:
|
I know my computer can choose DDDt to do, but my concern is not how long a work unit takes, is how long Boinc wastes. The reason why I periodically write that members who have slower machines and/or who switch them off very often should choose DDD-T is not because its WUs are shorter to run. This is already a good reason, but the most important is that DDD-T is currently the only project in WCG which can take so frequent checkpoints (one every time the percentage is a round integer, i.e. 100 times per WU). With these WUs one has no need to worry if it is a good time for switching off or not. Even with the slowest machines the lost time will never exceed 5 minutes and the job will progress session after session, which is not the case for any of the other projects if your sessions are really short. And if DDD-T is not the project you prefer this is not a problem either: WCG servers keep the balance between projects fair and if you receive more DDD-T WUs members who can crunch everything will receive more of the other projects, including your preferred one. Note that if you want to benefit from this particular "feature" you must keep the "write to disk at most every xx seconds" parameter to its default value of 60 seconds because checkpoints are not taken during this interval. Cheers. Jean. |
||
|
|
|