| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
| World Community Grid Forums
|
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 24
|
|
| Author |
|
|
knreed
Former World Community Grid Tech Joined: Nov 8, 2004 Post Count: 4504 Status: Offline Project Badges:
|
We have successfully completed our testing of the validator for non-redundant workunits for The Clean Energy Project. We have put the new validator into production and started loading workunits without redundancy. These new workunits should start being distributed in about 2 days.
There will be random workunits selected for double checking. We will also double check results sent to computers that have not yet demonstrated that they can reliably produce successful valid results. This change will accelerate the progress of the project and complete about 1.8 times as much research using the same computer power. Thank you for your ongoing support! |
||
|
|
gb009761
Master Cruncher Scotland Joined: Apr 6, 2005 Post Count: 3010 Status: Offline Project Badges:
|
Kevin, has there been any progress yet on upping/removing the 12Hr time limit?
----------------------------------------![]() |
||
|
|
Falconet
Master Cruncher Portugal Joined: Mar 9, 2009 Post Count: 3315 Status: Offline Project Badges:
|
Kevin, has there been any progress yet on upping/removing the 12Hr time limit? And maybe on checkpointing? ![]() - AMD Ryzen 5 1600AF 6C/12T 3.2 GHz - 85W - AMD Ryzen 5 2500U 4C/8T 2.0 GHz - 28W - AMD Ryzen 7 7730U 8C/16T 3.0 GHz |
||
|
|
knreed
Former World Community Grid Tech Joined: Nov 8, 2004 Post Count: 4504 Status: Offline Project Badges:
|
We are making some progress on the timeout changes.
We are not planning any changes to checkpointing at this time. The checkpointing changes would be significant and intrusive into the research application and require saving large sets of data frequently. It would make the application even more IO intensive than it is now. We do understand that the checkpoints are much further apart than we would like, but some things are out of our control due to the nature of the research. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Make the tasks checkpoint when shutting down, please :)
|
||
|
|
pball1224
Cruncher USA Joined: Nov 22, 2005 Post Count: 45 Status: Offline Project Badges:
|
That's sad news about checkpointing. Due to changes in my personal pc usage, I can no longer run CEP. I have been turning my machine off daily lately and have been seeing some tasks start from 0% over and over even after more than an hour of run time. :-(
|
||
|
|
anhhai
Veteran Cruncher Joined: Mar 22, 2005 Post Count: 839 Status: Offline Project Badges:
|
question for CA's or scientist:
----------------------------------------Currently I have 1 slow machine that crunch CEP2. It doesn't always finish step 2 (the critical one that scientist wants), but I thought of it as it doesn't matter because my WU will just be used to validate someone else's who has a faster machine. But now that we are doing zero redundancy, should I stop crunching CEP2 on that machine? ![]() |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Will the non-redundant work units carry the same nomenclature, or will they be identifiable? I.e., will they all be name_0, or will the trailing quorum indicator be gone entirely?
|
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hi anhhai,
I am not authoritative, but common sense says that a computer that slow should be switched to projects without a time limit. CEP2 is not the right project for your slow computer. Lawrence |
||
|
|
Rickjb
Veteran Cruncher Australia Joined: Sep 17, 2006 Post Count: 666 Status: Offline Project Badges:
|
@knreed: Congratulations on getting single-redundancy implemented. I'd been wondering when the results of the recent beta test would bear fruit.
----------------------------------------Pity about the checkpointing. A mod that looks like it would be not a huge task and might reduce disk I/O bottlenecks when multiple CEP2 WUs are run simultaneously, would be to merge into one directory the many common data files that are duplicated into each slot directory at WU startup time. Just a suggestion ... @pball1224: "I have been turning my machine off daily lately and have been seeing some tasks start from 0% over and over even after more than an hour of run time." If your computer is running Windows (2000 and later), you may be able hibernate it instead of shutting it down normally. Hibernate writes a complete image of the RAM onto a file in the root directory of the boot hard drive, and saves the status of all processes. The reverse happens at startup, and processes resume as if nothing had happened. Power can be completely turned off during hibernation, and it works for BOINC and CEP2 tasks. [Edit]: IIRC, it is wise to Suspend processng from the BOINC Manager Activity menu before hibernating, and of course remember to resume activity after starting up again. However, I've found that many machines won't hibernate, even though they are unning 2k, XP or 7 and hibernation is enabled; after selecting Hibernate and clicking OK on the Shut Down popup window, the screen goes black for a few seconds but then resumes normal activity. There are no diagnostics. This can happen if there is not enough space on the HDD for the RAM image, and also if there are devices installed whose drivers do not support hibernation, but in my cases the reason(s) remain mysterious. It's worth a try though. BTW, in Windows 7, "Hybrid Sleep" should also preserve and restore BOINC/CEP2 tasks. [Edit 2 times, last edit by Rickjb at Nov 3, 2011 4:12:24 PM] |
||
|
|
|