| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Support
Forum: BOINC Agent Support
Thread: Problem: Invalid Working Units in large numbers. Please help. |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 68
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I believe that XS_Fr3ak's problems are likely hardware based.  Gasp!! Do you realise what you are saying, ozylynx? We know XS's computers were all designed and built by Superman and they simply do NOT produce errors. That we know because Superman (aka Clark Kent aka Movieman) told Dagorath so in these very forums. The only way that machine could produce an error is if there was green kryptonite present when ol' Sup built it. I think this is Brainic5's handiwork. OK, there is one other explanation... an incognito Dutch Power Cow sold them that machine knowing it was junk in order to slow down XS's climb up the ratings. Nothing else makes any sense, nothing. Edit: Doh! How could I have missed it. Brainic5 is a Dutch Power Cow!!! That's it!!!! [Edit 1 times, last edit by Former Member at Nov 30, 2006 8:11:08 AM] |
||
|
|
depriens
Senior Cruncher The Netherlands Joined: Jul 29, 2005 Post Count: 350 Status: Offline Project Badges: 
|
I believe that XS_Fr3ak's problems are likely hardware based. Gasp!! Do you realise what you are saying, ozylynx? We know XS's computers were all designed and built by Superman and they simply do NOT produce errors. That we know because Superman (aka Clark Kent aka Movieman) told Dagorath so in these very forums. The only way that machine could produce an error is if there was green kryptonite present when ol' Sup built it. I think this is Brainic5's handiwork. OK, there is one other explanation... an incognito Dutch Power Cow sold them that machine knowing it was junk in order to slow down XS's climb up the ratings. Nothing else makes any sense, nothing. Edit: Doh! How could I have missed it. Brainic5 is a Dutch Power Cow!!! That's it!!!! Yes!! Can we welcome another new member??? Plz don't keep it for yourself and share it with us!      |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I believe that XS_Fr3ak's problems are likely hardware based. Gasp!! Do you realise what you are saying, ozylynx? We know XS's computers were all designed and built by Superman and they simply do NOT produce errors. That we know because Superman (aka Clark Kent aka Movieman) told Dagorath so in these very forums. The only way that machine could produce an error is if there was green kryptonite present when ol' Sup built it. I think this is Brainic5's handiwork. OK, there is one other explanation... an incognito Dutch Power Cow sold them that machine knowing it was junk in order to slow down XS's climb up the ratings. Nothing else makes any sense, nothing. Edit: Doh! How could I have missed it. Brainic5 is a Dutch Power Cow!!! That's it!!!! Thank youfor that post. I thought people on this forum were mature enough to have a unbiased view at the topic. Too bad I was wrong. I am not going to post anything again after this post. Of course our PCs are just as instable as any other PC on this planet, so they produce errors from time to time, but what a coincidence that a lot of those invalid claims appeared after the credit system was changed. And to one of the questions I asked earlier: It doesnt matter if BOINC uses SSE2, 3, 4 or whatever. The science part has to make use of it, no matter what boinc supports and what not. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
roan won't be replying. WCG chose to ignore Dagorath for a while, since he wasn't causing trouble. But clearly, he has started to be insulting again.
XS_Fr3ak, you are attempting to see a coincidence where no coincidence exists. The WCG science applications have been upgraded several times recently, and it is far more likely to be one of these changes causing an issue on a couple of your computers. There is certainly no conspiracy! What purpose would that serve? |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
And to one of the questions I asked earlier: It doesnt matter if BOINC uses SSE2, 3, 4 or whatever. The science part has to make use of it, no matter what boinc supports and what not. Wrong. The science app is not required to use SSE at all. If the app is written and compiled to use SSE then it will. If it is not written and compiled to use SSE then it will not. -- This post has been edited to comply with forum rules - nelsoc [Edit 1 times, last edit by Former Member at Dec 1, 2006 1:37:41 AM] |
||
|
|
keithhenry
Ace Cruncher Senile old farts of the world ....uh.....uh..... nevermind Joined: Nov 18, 2004 Post Count: 18667 Status: Offline Project Badges: 
|
Yes, let's keep this repectful folks. We're all working for the same goal though our motivations may vary.
----------------------------------------Just a thought but if there is a problem with a few machines yeilding a large number of invalid now that weren't prior to the change, what is it about the change that is catching them now that didn't previously. This may be a pre-existing problem that is being caught now. From what little I understand, if the science returned is bad, the WU would get flagged as an error. If the WU is being flagged as invalid purely because of the outlying credit claim, it seems to still be included as one of the quorum of 3 results. If the WU gets resent to a 4th, then it was invalid for an entirely different reason. The health checks would serve to at least rule out the possibility of hardware problems. If that is the case, then it turns the focus more to what is happening on the WCG servers. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I am following up on the problem originally posted on Nov 23rd in this thread. My original post was edited on Nov 28th to provide the Computer ID information as requested by nelsoc and uplinger of WCG project.
Sadly, no official feedback has been provided up to this time. The source of the problem has not been identified yet. We are at a loss of what to do. We strongly believe that a serious problem exists but we need help from the WCG administration to pin point the source and save all those wasted computer cycles. Unfortunately, while we are still waiting, valuable computer time is being lost and at an alarming rate. For example, just today it was reported by one of our members (member name : ddtung) that he has two full pages of new Invalid results from ten different machines. I hope that WCG admins are able to inspect those Invalid WUs and trace the source of the problem. During the past two weeks since the problem was brought to your attention, a number of community advisors have tried to offer their assistance by posting valuable suggestions. We are grateful, and appreciate their effort. However non of those suggestions would solve the problem. Please keep in mind that as valuable those recommendations are, they are the first things we look at when something like this happens. The machines exhibiting those problems are run by members who are well aware of those recommendations and have been running WUs for this project (and many others). The problem is new, not the machines. Only WCG admins have access to the other returned results, and have the knowledge needed to analyse the results log files for signs of the problems. So we need your help in doing so. Allow me to convey the high level of frustration of the team members for not being able to solve this problem. Please help us help you. We are open to all suggestions. Urgently awaiting your feedback. Thank you very much meshmesh On behalf of team XtremeSystems PS: Member ddtung is by no means alone. Here are a few more examples from other members as well: x2 4400 @ 2.2 2x1ghz ram XP pro Computer ID# 65518 faah0959_ d085n266_ x1AJX_ 00 Valid 11/28/2006 04:31:20 11/28/2006 17:48:58 10.85 78/63 faah0959_ d085n266_ x1AJX_ 00 Invalid 11/27/2006 21:48:54 11/28/2006 04:27:09 6.34 79/32 <== faah0959_ d085n266_ x1AJX_ 00 Valid 11/20/2006 22:01:37 11/22/2006 03:20:18 17.24 54/63 faah0959_ d085n266_ x1AJX_ 00 Error 11/20/2006 21:48:56 11/20/2006 21:54:05 0.00 0/0 faah0959_ d085n266_ x1AJX_ 00 Valid 11/20/2006 21:44:05 11/28/2006 13:35:10 3.57 73/63 faah0959_ d085n266_ x1AJX_ 00 Valid 11/20/2006 21:35:48 11/21/2006 10:09:01 6.96 62/63 Computer ID# 69672 faah0991_ d149n705_ x1AJX_ 01 Invalid 11/26/2006 16:09:15 11/29/2006 01:56:54 4.00 76/44 <=== AMD X2 3800, Computer ID# 72463 B10615_ 0337_ CTMA3A2-6-9-24 Valid 12/01/2006 16:23:54 12/02/2006 15:24:22 5.40 46/47 B10615_ 0337_ CTMA3A2-6-9-24 Valid 11/30/2006 08:52:05 12/01/2006 16:19:48 8.82 56/47 B10615_ 0337_ CTMA3A2-6-9-24 Invalid 11/30/2006 07:57:11 11/30/2006 19:24:45 7.13 86/24 <== B10615_ 0337_ CTMA3A2-6-9-24 Valid 11/30/2006 07:54:19 12/01/2006 00:21:29 6.11 39/47 Pentium D 820 @ 3.4, 1GB , Client 5.4.11, Computer ID# 61895. B10607_ 0065_ CTMA3A1-14-11-4-c1 Valid 12/01/2006 09:49:11 12/01/2006 23:55:04 5.15 33/40 B10607_ 0065_ CTMA3A1-14-11-4-c1 Valid 11/30/2006 22:55:18 12/01/2006 08:39:31 3.12 43/40 B10607_ 0065_ CTMA3A1-14-11-4-c1 Invalid 11/28/2006 13:30:08 11/30/2006 20:49:19 8.26 72/20 <== B10607_ 0065_ CTMA3A1-14-11-4-c1 Invalid 11/28/2006 13:30:00 11/30/2006 21:54:08 8.70 80/20 B10607_ 0065_ CTMA3A1-14-11-4-c1 Valid 11/28/2006 13:25:41 11/30/2006 07:16:20 3.59 43/40 x2 4400 @ 2.2 2x1ghz ram XP pro, Computer Id# 65518 faah0998_ d162n041_ x1AJX_ 00 Valid 12/01/2006 04:51:45 12/02/2006 12:44:43 17.43 86/80 faah0998_ d162n041_ x1AJX_ 00 Valid 11/29/2006 13:50:23 11/30/2006 00:31:21 5.12 81/80 faah0998_ d162n041_ x1AJX_ 00 Invalid 11/29/2006 13:50:23 12/01/2006 04:44:45 6.44 80/40 <== faah0998_ d162n041_ x1AJX_ 00 Valid 11/29/2006 13:39:37 11/30/2006 13:46:12 9.15 75/80 Username: joshd, host name: josh B05558_ 0131_ CTMA4A-36-2-17-c2 Invalid 11/22/2006 03:44:33 11/30/2006 22:11:15 3.97 25/8 <== B05526_ 0067_ CTMA3A2-17-15-10-c2 Invalid 11/18/2006 09:10:25 11/27/2006 02:31:50 4.32 27/12 <== |
||
|
|
uplinger
Former World Community Grid Tech Joined: May 23, 2005 Post Count: 3952 Status: Offline Project Badges: 
|
Hello,
I have looked into the hosts that you have shown. host id 65518: Has an invalid workunit on a beta workunit. The issue with it being invalid is that we were having mis compares between AMD an Intel processors. This has been fixed in the latest version. host id 69672: Has on client compute error out of 100 workunits. The workunit he errored on was completed by others so it was probably an issue with accessing too much memory when the machine had allocated it to other applications. host id 72463: I looked to see if i could find any invalids posted by this member in the past few days and did not see the one in reference. The reason I speculate though is that we had an issue with the way BOINC points were being checked. If a member claimed too high then it was showing up as invalid. This has been fixed as of 2 weeks ago. host id 61895: Same as host 72463 joshd hostname josh: I do not see those workunits in the database for him. His points seem proper, but I can not know without knowing what other received for those workunits. -Uplinger |
||
|
|
|