| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: Say No to Schistosoma Forum
Thread: [resolved] general problem with vina based projects |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 6
|
|
| Author |
|
|
sgoll
Advanced Cruncher Joined: Oct 24, 2006 Post Count: 87 Status: Offline Project Badges: 
|
Right now there are three active vina based projects, the DSFL, GFAM and SN2S. I have three boxes, two are AMD geode LX driven and one Via C3 Nehemiah. All were running on other project before with nearly no errors so the boxes are fiine. Nothing overclocked, no memory shortage, same setup as my other boxes. The only difference is that these three boxes does not have one of the common CPUs (historical seen they are all cyrix based) and the have hardware random number generators. The operating system is a 32 bit debian linux (as one my other boxes), they run the latest 5.10.58 boinc client, the geodes are 586 class systems and the Nehemiah is fully 686 compliant. They all have the 686 class cmov instruction.
----------------------------------------The logs are ending like this one: /www.worldcommunitygrid.org/sn2s.target_AAC62255.pdbqt size = 109 0 [20:32:50] Finished task #34 cpu time used 13328.750000 [20:32:50] Starting task 35,CPU time is 722007.590000. [20:32:50] ./ZINC08848952.pdbqt size = 27 8 ../../projects/www.worldcommunitygrid.org/sn2s.target_AAC62255.pdbqt size = 109 0 [00:34:22] Finished task #35 cpu time used 13136.790000 00:34:23 (20939): called boinc_finish </stderr_txt> ]]> or like this: /www.worldcommunitygrid.org/sn2s.target_AAF21679.pdbqt size = 342 0 [07:31:07] Finished task #70 cpu time used 6691.746208 [07:31:07] Starting task 71,CPU time is 223024.986162. [07:31:07] ./ZINC25250150.pdbqt size = 33 9 ../../projects/www.worldcommunitygrid.org/sn2s.target_AAF21679.pdbqt size = 342 0 [09:24:46] Finished task #71 cpu time used 6684.805774 09:24:46 (5090): called boinc_finish </stderr_txt> ]]> Right now they are running Say No to Schistosoma because I wanted to make sure that there is no memoroy shortage and the boxes have enough memory for cache and buffers. No sign of an error, but all results on these three boxes are validated as invalid. When switching to FAAH or HPF2 they are generating valid results. Well: that's it. For me it looks like a problem with the vina code, but I'm aware that I'm the code is running fine on more actual CPUs. May be there is a problem with the validators (these systems running up to 3 ... 8 days on one workunit) but this "should" not have an influence. SN2S_ AAC62255_ 0000127_ 0749_ 0-- alix02 Invalid 09/10/12 22:02:04 19/10/12 23:34:49 204.21 286.5 / 128.8 I have no idea what could be the cause for the problems and I know that these three boxes are very uncommon "crunching" systems. But I'm curious and that's why I'm asking: does anyone have an idea? :) Stephan  [Edit 2 times, last edit by sgoll at Oct 24, 2012 11:34:42 AM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
AMD and Vina do not agree, they will only give valid results when your wingman has the same family of cpu. I also have 2 family 6 boxes, they run the wu fine but only go valid 1 out of about 50 times. It has been decided that as there are only a few of us with these older CPUs that nothing will be done. Any other (non-Vina) project is fine
|
||
|
|
Falconet
Master Cruncher Portugal Joined: Mar 9, 2009 Post Count: 3315 Status: Offline Project Badges: 
|
I have an athlon II X2 445 which is also from family 6 and so far it hasn't gotten any invalids on any science. In GFAM it is even with single redundancy.
---------------------------------------- - 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 |
||
|
|
KWSN - A Shrubbery
Master Cruncher Joined: Jan 8, 2006 Post Count: 1585 Status: Offline |
AMD and Vina do not agree, they will only give valid results when your wingman has the same family of cpu. I also have 2 family 6 boxes, they run the wu fine but only go valid 1 out of about 50 times. It has been decided that as there are only a few of us with these older CPUs that nothing will be done. Any other (non-Vina) project is fine Fredski, this is a common mis-perception that needs to stop being repeated. AMD has absolutely no issue with Vina applications. Older AMD cpus - as in XP class will not validate unless paired with a similar CPU. For the OP, the techs are well aware of this issue and have stated that this class of CPUs does not contain enough computers to justify a separate redundancy rule. The only solution is to uncheck those sciences from your profile for the systems producing invalid results.  Distributed computing volunteer since September 27, 2000 |
||
|
|
sgoll
Advanced Cruncher Joined: Oct 24, 2006 Post Count: 87 Status: Offline Project Badges:
|
... Older AMD cpus - as in XP class will not validate unless paired with a similar CPU. For the OP, the techs are well aware of this issue and have stated that this class of CPUs does not contain enough computers to justify a separate redundancy rule. The only solution is to uncheck those sciences from your profile for the systems producing invalid results. Well, please allow me to correct you. The AMD XP class systems should validate fine. GFAM_ x2FFCa_ 6WATs_ PvODCase_ 0045935_ 0135_ 1-- gunboat Valid 22/10/12 10:53:46 24/10/12 08:23:02 12.48 179.9 / 224.4 DSFL_ 00020-45_ 0000047_ 0170_ 0-- gunboat Valid 22/10/12 02:39:30 23/10/12 21:01:11 8.20 108.4 / 108.4 SN2S_ AAF21681_ 0000257_ 0655_ 0-- gunboat Valid 21/10/12 16:12:55 23/10/12 21:01:11 8.24 110.3 / 110.3 DSFL_ 00020-44_ 0000037_ 0623_ 0-- gunboat Valid 21/10/12 04:43:47 23/10/12 08:36:58 5.93 76.6 / 76.6 GFAM_ xMRC_ CMxtl257_ wWATs_ PfSUB1_ 0045010_ 0110_ 1-- gunboat Valid 19/10/12 15:12:18 21/10/12 20:25:06 10.54 211.9 / 228.9 This is on my K7 Athlon XP-MP 2800+ (basically an XP with guaranteed multiprocessor capabilities) driven linux system. You can see that all three vina projects are fine on this system. I also have AMD Athlons (also K7class), but they are not crunching, but my XP-MP did not generate more invalid results than any other system. But it is nice to read that there is a known vina issue on older CPUs. This is what I wantet to know when I asked. I will add an "resolved" flag to the header. Thanks, Stephan |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
AMD and Vina do not agree, they will only give valid results when your wingman has the same family of cpu. I also have 2 family 6 boxes, they run the wu fine but only go valid 1 out of about 50 times. It has been decided that as there are only a few of us with these older CPUs that nothing will be done. Any other (non-Vina) project is fine Fredski, this is a common mis-perception that needs to stop being repeated. AMD has absolutely no issue with Vina applications. Older AMD cpus - as in XP class will not validate unless paired with a similar CPU. For the OP, the techs are well aware of this issue and have stated that this class of CPUs does not contain enough computers to justify a separate redundancy rule. The only solution is to uncheck those sciences from your profile for the systems producing invalid results. Just to show it is NOT just older AMD that have this problem, it is just the older AMD that do not have a work around in place (nor should they) https://secure.worldcommunitygrid.org/forums/...ead,32982_offset,0#373265 |
||
|
|
|