| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Beta Testing
Forum: Beta Test Support Forum
Thread: New Beta starting Aug 23, 2011 |
| No member browsing this thread |
|
Thread Status: Locked Total posts in this thread: 164
|
|
| Author |
|
|
littlepeaks
Veteran Cruncher USA Joined: Apr 28, 2007 Post Count: 748 Status: Offline Project Badges: 
|
I received a few, and they're starting to crunch. It is estimating total time to completion as 7:52:06, but appears they will complete in well under 6 hours.
As usual, there's no graphics with these WUs, but I LOVE the background color on the graphics window, especially if you make it full screen.  I have a P7 with Windows 7 64 bit. |
||
|
|
Dataman
Ace Cruncher Joined: Nov 16, 2004 Post Count: 4865 Status: Offline Project Badges: 
|
Looks like we are done. Nice run.
----------------------------------------  |
||
|
|
KWSN - A Shrubbery
Master Cruncher Joined: Jan 8, 2006 Post Count: 1585 Status: Offline |
My Windows XP machine became very, very unresponsive with this beta running. It cleared up with a reboot. Is there a memory leak going on here possibly?
----------------------------------------On the bright side, the reboot didn't affect the progress of the work unit. It restarted from the checkpoint without any difficulty. Also, it looks like you'll need to up the number of computations or I got a really unlucky streak of smaller units. Nothing I had ran for over six hours on the slowest of systems. My Pentium III Edit: The XP machine became unresponsive again within a couple minutes of rebooting. BOINC service continued in the background and as soon as the Beta unit completed the computer was accessible. Looks like it's not running as a background priority, at least in Win XP 32 bit SP3. As for the PIII result, I read the wrong line in my results. Looks like this one will take just under 12 hours. That's still far below the size needed for an average computer to last six hours on a unit.  Distributed computing volunteer since September 27, 2000 [Edit 1 times, last edit by KWSN - A Shrubbery at Aug 24, 2011 5:07:43 AM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
My Windows XP machine became very, very unresponsive with this beta running. It cleared up with a reboot. Is there a memory leak going on here possibly? On the bright side, the reboot didn't affect the progress of the work unit. It restarted from the checkpoint without any difficulty. Also, it looks like you'll need to up the number of computations or I got a really unlucky streak of smaller units. Nothing I had ran for over six hours on the slowest of systems. My Pentium III Edit: The XP machine became unresponsive again within a couple minutes of rebooting. BOINC service continued in the background and as soon as the Beta unit completed the computer was accessible. Looks like it's not running as a background priority, at least in Win XP 32 bit SP3. As for the PIII result, I read the wrong line in my results. Looks like this one will take just under 12 hours. That's still far below the size needed for an average computer to last six hours on a unit. Found similar issue, when snoozing wu or using cpu, take 30-40 sec for wu to stop or release cpu cycles, not a big issue for me as am used to waiting for the creeper to respond ;) |
||
|
|
kskjold
Senior Cruncher Norway Joined: May 20, 2008 Post Count: 469 Status: Offline Project Badges: 
|
Four received and returned and now in PV. On Win 7 I7 850 8GB ram. I was a sleep when they where running, but it doesn't seem to have been any problems with them. The ran for a little bit over 4.6 hours.
---------------------------------------- |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
As I previously reported, on several XP machines (from P4-HT to C2D) they work OK with ONE WU at a time. They seem to obey CPU time limits and snooze command as expected.
Now, having said that, I do find some strange things, which I also reported on the previous beta run: 1) In some cases there's a huge difference between Elapsed Time and CPU Time. 2) Even if apparently this WUs obey CPU time limits set in BOINC, the processor temperature is several °C higher. From a couple to 6 or more. There must be something different in the way these WUs run compared to all other projects. I don't know what happens when there are 2 or more simultaneous WUs. In the earlier run it did cause me problems. |
||
|
|
Crystal Pellet
Veteran Cruncher Joined: May 21, 2008 Post Count: 1322 Status: Offline Project Badges: 
|
I woke up this morning, BETA was on my side:
----------------------------------------Most important part of my report: Suspending a 6.14 task on XP32 @1500MHz needs 3 seconds to reach 0% CPU. -------------------------------- Cleanup the 6.13's: Keith doesn't care about the results of the inadvertently sent 6.13's, but perhaps about the cleanup. In the morning I first looked in BoincTasks and saw 3 BETA's 'Ready to start'. Pushed them up in queue and they started. Then I read 6.13's could be aborted, they were. So I did! I see that after a while from the aborts resends are sent. I suppose 6.14's, but can't see. [Edit 1 times, last edit by Crystal Pellet at Aug 24, 2011 7:05:53 AM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Received one 6.14 on W7-32. No issues. Suspended the job just after push-start and at 5 CPU hours. Prompt stop of the worker app and resume, not time losses. Run full tilt, 2C up, but that's not unusual as different sciences use varying functions more intensely. Computer, duo, remains fully responsive.
Now it's checking the result log after completion to see how much the last step Wallclock v. CPU time is. Given the speed of the computer, it's heading to complete at 7 hours, which is not bad at all. It's an ampc type job btw. @uplinger, your explanation was what was understood. Dynamically tied job size cutting towards the actual mean computing time of the current contributors to the science, I.e. would a fleet of P3 suddenly enter the crunch, their added mean computing time would tell the sizer, to put less ligands in, to have the average for the project stay at ~ 6 hours, the variation on the theme. I'd expect if this would be implemented at host level, you'd have to consider matching CPU's as with HCMD2, [quorum 2]. Curious if the plan is to run this "Vina engined" project for the full duration in quorum 2 or to later switch to zero redundancy, similar to FAAH and HFCC. --//-- |
||
|
|
marvey11
Advanced Cruncher Germany Joined: Apr 2, 2011 Post Count: 89 Status: Offline Project Badges: 
|
Got a result for the workunit BETA_BETA_ampc_0000001_0917 (40 internal tasks).
----------------------------------------My system: Linux, Intel Pentium-M 1600 MHz, BOINC 6.12.33 * no problems with task suspension * system responsiveness OK * temperature is 63 to 64 °C which is slightly higher than average (around 60 to 61 °C) but still about the same as with HCC tasks, so no problem here either * looks like the final CPU time will be between around 5 and 5.25 hours, so I concur that the task sizes could probably be increased since I would consider this machine below average Something else I did notice, though, is an unusually high number of science-app subprocesses. The usual is 3 processes as can be seen with the suspended HCC task below. The beta task has 6 of these subprocesses, however. Since I have no clue whether this is what it's supposed to be I thought I'd rather mention it. marvey@helium:~$ ps -u boinc -U boinc -o pid,ppid,comm  |
||
|
|
Coleslaw
Veteran Cruncher USA Joined: Mar 29, 2007 Post Count: 1343 Status: Offline Project Badges:
|
So far, all other WU's are doing fine on my systems. Just that one that errored.
----------------------------------------    |
||
|
|
|