| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: Help Defeat Cancer
Thread: Recovery from checkpoints does not always work |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 17
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I have found (trial and error) that if you "Exit" BOINC/HDC before you shutdown / reboot / standby your system, then it will restart from the previous checkpoint. If you don't, then it will very often restart from zero. Not a good thing.
This is something useful to know, but it's only a work-around. IMO this should be an install-and-forget application. Users shouldn't have to remember to close down BOINC before rebooting their machine. I don't know if it's only an issue with HDC or with the other WCG projects as well, but I believe it should be fixed at some point. |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
barbanis,
----------------------------------------it's a good point. Thing is, many users have forced their Windows to not wait for proper closing of their apps and force exit before the state files have been written to disk. There are ways to delay the close of windows to give it time to do so. Keep in mind, that the HDC's computation files are often 1gb in size.... that requires time to write to disk. Members and other CA's keep on repeating the 'Hibernation ' function . Snooze BOiNC or UD Agent, proceed to hit the start, exit button of windows and select the option. It writes the complete memory state to disk. Upon return, the machine will resume as if it was never off. It's rescuing me regularly when the power fails and the UPS starts the clsoing procedure after 5 minutes. Set and forget would be ideal, but the ideal world, i fear can only be found at Nirvana. cheers
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hmm, seems HDC is mis-behaving more and more. Last time, I snoozed BOINC, then exited it. When I restarted it, I got this message:
12/5/2006 9:32:55 AM|World Community Grid|Unrecoverable error for result B10627_0230_CTMA3A2-13-23-21_0 ( - exit code -1073741819 (0xc0000005)) Dunno... seems HDC wants to run uninterrupted from start to finish, or else. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I'll second that. It really doesn't like being disturbed....
Jonathan. |
||
|
|
armstrdj
Former World Community Grid Tech Joined: Oct 21, 2004 Post Count: 695 Status: Offline Project Badges: 
|
There is a problem with checkpointing on some of the new workunits. A fix is being tested and should be available soon.
-armstrdj |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Great news! thanks very much!
 |
||
|
|
Kaoh
Cruncher ROC,Taiwan(NOT PRC) Joined: Nov 20, 2005 Post Count: 18 Status: Offline Project Badges: 
|
Seems like I meet problem again...
----------------------------------------from 50% to 0%... I'm sure I wait it fully"re-started", when it began to proceed,I closed BOINC... Now the issue is found,wait and see...  [Edit 1 times, last edit by Kaoh at Dec 6, 2006 4:25:04 PM] |
||
|
|
|