| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: Help Defeat Cancer
Thread: CPU/Disk Load to 100% every 15 minutes |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 29
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I have the very same problem: every 15 minutes, disk drive goes wild. Problem is, once I/O starts, it's difficult to get anything else done. So, for 3-5 minutes, I better not need to execute any I/O.
It seems these checkpoints aren't taken after the task is >30 to 50% completed. As somebody suggested, I installed BOINC. But, unfortunately, it does not run HDC. As this problem has the annoying tendency to manifest itself when I absolutely need the PC to be responsive, I had no other choice but to exclude the project from my profile. At least until this behaviour is fixed. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
BOINC does run HDC. I don't follow what you are saying.
If you need help setting it up and attaching to WCG, please ask. We're happy to help. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I had no other choice but to exclude the project from my profile. Which project are you running in BOINC? What are your computer specs? I am running all projects - BOINC on one machine and UD on the other. No problems. Lawrence |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
When I installed BOINC, I gave it a profile to run only HDC. It remained idle until I added other projects.
Also, from the BOINC - Getting Started web page @ http://www.worldcommunitygrid.org/help/viewTopic.do?shortName=bnc "What projects at World Community Grid are available on BOINC? The Human Proteome Folding project and FightAIDS@Home are currently available on BOINC for Windows and Linux. Only FightAIDS@Home is available for Mac." Maybe this entry is obsolete, but the fact is that BOINC remained idle when it had only HDC in its profile. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hello CSA,
The current projects available are HDC (750 MB RAM) and FAAH (250 MB RAM). Since there are more computers available to run HDC than there are work units generated each day, people selecting HDC only should also check the box to run other projects if no work units are available. BOINC HDC should not cause much Disk I/O unless a memory shortage causes a lot of page faulting in Virtual Memory. Could this be your problem? HPF2 is currently being debugged before reactivation ( http://www.worldcommunitygrid.org/forums/wcg/viewthread?thread=8062 ). Lawrence |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
It's always 'good' to read at other language forums of other projects, that WCG is not alone with intense disk i/o phenomena....what i'm saying is, everywhere the models are getting more intense and a balance needs to be found how often a work unit checkpoints back to disk.......even with 1.5gb HDC at times is causing little delays, but only when i'm accessing the disk at the same time.....running HDC on dual core, 2 simultaneous will be stressful.....option exists to reduce BOINC use to 1 core, and now also throttle it down like the UD agent already can (from 5.6.x not yet supported !!!!!!).
----------------------------------------@CSA, if on BOINC, there is the 'Alternate Work' Option for each device at bottom of projects selection (few months now and BOINC only) .......just got 7 FAAH because of this message: 06-09-20 8:38:28|World Community Grid|Message from server: No work can be sent for the projects you have selected 06-09-20 8:38:28|World Community Grid|Message from server: You have selected to receive work from other projects if no work is available for the projects you selected 06-09-20 8:38:28|World Community Grid|Message from server: Sending work from other projects ciao
WCG
----------------------------------------Please help to make the Forums an enjoyable experience for All! [Edit 1 times, last edit by Sekerob at Sep 20, 2006 6:49:42 AM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Apparently there simply wasn't any HDC work available during my first try. I tried once again, and BOINC is running HDC work. And, thankfuly, BOINC does not monopolize the PC with disk I/O.
Thanks for the feedback!  Carlos Montreal, Canada |
|||
|
||||
|
Richard L Lozes
Cruncher Joined: Jun 21, 2006 Post Count: 6 Status: Offline |
Same problem here. Disk I/O is excessive for HDC. I've also got 1GB DRAM, a 2 GB page file (fixed, contiguous allocation), with plenty of left-over space in both. In ~10 minutes since rebooting, wcg_HDC_TMA.exe has accumulated 827k page faults.
The situation is not helped by SpySweeper also accumulating 875k page faults in the same period. I'm not yet sure, but it seems to me that one begets the other. It would be helpful for the authors to recode their application; I don't mind contributing unobtrusive cycles (i.e., low priority CPU), but this disk thrashing gets in the way of my use of the machine. Furthermore, I guarantee the task can get significantly more results per unit time - thrashing drags the CPU utilization down almost to 0%. |
|||
|
||||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
A conflict? That's a strong possibility. Badly designed antivirus/spyware programs often interfere with low-priority processes like WCG.
Try disabling any continuous scanning on the security software. That may identify it as the definite problem. It's amazing what garbage you can get away with by yelling "security! danger!" and hawking a shoddy solution - it's a tactic that has been tried by not a few world leaders as well.... |
||
|
|
|