| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Active Research
Forum: OpenPandemics - COVID-19 Project
Thread: Invalid GPU work units |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 171
|
|
| Author |
|
|
bluestang
Senior Cruncher USA Joined: Oct 1, 2010 Post Count: 272 Status: Offline Project Badges: 
|
Up to 16 pages (about 240 WUs) of Invalids for me.
----------------------------------------Posting just for reference as you guys are working to solve the issue. |
||
|
|
nanoprobe
Master Cruncher Classified Joined: Aug 29, 2008 Post Count: 2998 Status: Offline Project Badges: 
|
I’m at work with no way to remote in and abort them sooo.....
----------------------------------------Hopefully the ones that get processed will get some credit for the time/effort. 🙃
In 1969 I took an oath to defend and protect the U S Constitution against all enemies, both foreign and Domestic. There was no expiration date.
  |
||
|
|
Grumpy Swede
Master Cruncher Svíþjóð Joined: Apr 10, 2020 Post Count: 2165 Status: Offline Project Badges: 
|
Hmm, got 3 from the 2573 batch, and they validate fine. So, some WU's are OK, even from the "faulty" batches.
----------------------------------------Edit: And Now I just received a bunch of them from batches: 2636, and 2670. Let's see how that goes. So, obviously batches 2225-3336 are not stopped from going out. [Edit 1 times, last edit by Grumpy Swede at Apr 15, 2021 3:05:57 PM] |
||
|
|
pututu
Senior Cruncher United States Joined: Jan 3, 2016 Post Count: 241 Status: Offline Project Badges: 
|
Just a note, the newly created workunits like 2573, 2670, 2659 batch also validate fine on my hosts but appear to have lower credit score than before.
---------------------------------------- |
||
|
|
Grumpy Swede
Master Cruncher Svíþjóð Joined: Apr 10, 2020 Post Count: 2165 Status: Offline Project Badges:
|
Just got OPNG_0002410_00095_4, with two invalids already. That one is going to die a painful death, I'm sure of it
---------------------------------------- Edit: This one died a painful death: https://www.worldcommunitygrid.org/ms/device/...s.do?workunitId=621179514 And this one too: https://www.worldcommunitygrid.org/ms/device/...s.do?workunitId=621047195 Most of the other from the faulty batches became valid, low credit and very short runtime. [Edit 2 times, last edit by Grumpy Swede at Apr 15, 2021 3:41:19 PM] |
||
|
|
uplinger
Former World Community Grid Tech Joined: May 23, 2005 Post Count: 3952 Status: Offline Project Badges: 
|
Good morning,
I should have clarified more on my post last night. I have suspended all new work from those batches, however we have a mechanism that attempts a rerun of error work units. I have to manually catch these and change them. I did that for some of them last night, but during sleep some may have been rerun. I will work on catching more today. Also, something to note is that not all the work units become error from those batches. Thanks, -Uplinger |
||
|
|
nanoprobe
Master Cruncher Classified Joined: Aug 29, 2008 Post Count: 2998 Status: Offline Project Badges:
|
Since the tasks are finished so fast I tried and experiment. I received 8 on one of my R9-280x GPUs so I set it to process all 8 at the same time. All 8 completed in less than 2 1/2 minutes. That was fun. And FWIW it appears that the CPU processing was 65% of the total runtime. All 8 valid.
----------------------------------------
In 1969 I took an oath to defend and protect the U S Constitution against all enemies, both foreign and Domestic. There was no expiration date.
----------------------------------------[Edit 1 times, last edit by nanoprobe at Apr 15, 2021 9:00:41 PM] |
||
|
|
Grumpy Swede
Master Cruncher Svíþjóð Joined: Apr 10, 2020 Post Count: 2165 Status: Offline Project Badges:
|
Still rolling in new _0 tasks from those batches. A whole bunch of 2829 just recently. as well as from batch 2652. So. I still don't understand that they could be stopped from being sent out. These are not resends at all, but new WU's from the supposedly stopped batches.
----------------------------------------And very short runtimes, and very low credit. Not that credit matters much, but runtimes of only 40-50 seconds on a GTX980. That must be close to 10 seconds on a newer Nvidia GPU. Edit: And it's not just batches for OPNG 2225-3336 that have this problem. Just got one_4 from batch 3465 , and it too have lots of invalids: https://www.worldcommunitygrid.org/ms/device/...s.do?workunitId=621659533 [Edit 1 times, last edit by Grumpy Swede at Apr 15, 2021 11:40:39 PM] |
||
|
|
nanoprobe
Master Cruncher Classified Joined: Aug 29, 2008 Post Count: 2998 Status: Offline Project Badges:
|
Still rolling in new _0 tasks from those batches. A whole bunch of 2829 just recently. as well as from batch 2652. So. I still don't understand that they could be stopped from being sent out. These are not resends at all, but new WU's from the supposedly stopped batches. And very short runtimes, and very low credit. Not that credit matters much, but runtimes of only 40-50 seconds on a GTX980. That must be close to 10 seconds on a newer Nvidia GPU. Did the new _0 tasks end up valid or invalid? I had a bunch of _0 and _1 end up invalid.
In 1969 I took an oath to defend and protect the U S Constitution against all enemies, both foreign and Domestic. There was no expiration date.
----------------------------------------[Edit 1 times, last edit by nanoprobe at Apr 15, 2021 11:39:30 PM] |
||
|
|
Grumpy Swede
Master Cruncher Svíþjóð Joined: Apr 10, 2020 Post Count: 2165 Status: Offline Project Badges:
|
Still rolling in new _0 tasks from those batches. A whole bunch of 2829 just recently. as well as from batch 2652. So. I still don't understand that they could be stopped from being sent out. These are not resends at all, but new WU's from the supposedly stopped batches. And very short runtimes, and very low credit. Not that credit matters much, but runtimes of only 40-50 seconds on a GTX980. That must be close to 10 seconds on a newer Nvidia GPU. Did the new _0 tasks end up valid or invalid? I had a bunch of _0 and _1 end up invalid. For me they ended up as valid, but as you can see from my edit above, we now also have invalids from higher batches than what Uplinger said was the range 2225-3336 Edit: OPNG_0002701 _00456 also new, and now also going to invalids: https://www.worldcommunitygrid.org/ms/device/...s.do?workunitId=621682520 So, pretty clear by now that these batches are not stopped from sending out new WU's, and also clear that the upper limit of the faulty batches is not 3336. [Edit 1 times, last edit by Grumpy Swede at Apr 15, 2021 11:59:06 PM] |
||
|
|
|