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 5, 2011 |
No member browsing this thread |
Thread Status: Active Total posts in this thread: 94
|
Author |
|
kateiacy
Veteran Cruncher USA Joined: Jan 23, 2010 Post Count: 1027 Status: Offline Project Badges: |
Here are the details of my Phenom II X4 running 64-bit Ubuntu 10.04.3 (previously summarized in a table):
----------------------------------------BETA_ BETA20_ ace_ 0000000_ 0172_ 0-- kate-amd64 Invalid 8/5/11 18:09:02 8/5/11 23:04:06 3.27 86.7 / 59.6 BETA_ BETA20_ ace_ 0000000_ 0089_ 0-- kate-amd64 Valid 8/5/11 18:09:02 8/6/11 04:53:14 3.66 97.1 / 92.9 BETA_ BETA20_ ace_ 0000000_ 0080_ 0-- kate-amd64 Inconclusive 8/5/11 18:09:02 8/6/11 04:53:14 3.24 86.0 / 0.0 BETA_ BETA20_ ace_ 0000000_ 0782_ 1-- kate-amd64 Valid 8/5/11 18:09:02 8/5/11 21:16:54 3.04 81.1 / 83.3 BETA_ BETA_ ace80_ 0000000_ 4955_ 0-- kate-amd64 Valid 7/29/11 23:21:33 7/30/11 01:38:47 2.00 51.9 / 40.3 BETA_ BETA_ ace80_ 0000000_ 3837_ 0-- kate-amd64 Invalid 7/29/11 22:09:25 7/30/11 00:29:32 2.00 52.3 / 23.3 BETA_ BETA_ ace80_ 0000000_ 1737_ 1-- kate-amd64 Invalid 7/29/11 21:09:06 7/30/11 00:07:22 2.02 52.1 / 50.0 BETA_ BETA_ ace80_ 0000000_ 1607_ 0-- kate-amd64 Invalid 7/29/11 21:08:45 7/30/11 00:08:30 2.43 62.7 / 42.4 BETA_ BETA_ ace80_ 0000000_ 0448_ 1-- kate-amd64 Valid 7/29/11 20:27:05 7/29/11 23:21:33 2.38 61.8 / 63.0 BETA_ BETA_ ace80_ 0000000_ 0628_ 1-- kate-amd64 Valid 7/29/11 19:49:37 7/29/11 22:09:24 1.91 50.5 / 50.5 BETA_ BETA_ ace80_ 0000000_ 3560_ 1-- kate-amd64 Invalid 7/29/11 19:31:06 7/29/11 20:57:47 1.24 30.6 / 21.5 Not quite as grim as KerSamson's, but still 55% invalid (assuming the inconclusive turns out badly). This machine crunches WCG 24/7 without other errors or invalids (except on 1 particular batch of DDDT2). Out of curiosity, maybe I'll upgrade to Ubuntu 11.04 this weekend and see whether that makes any difference on next week's betas. (Probably won't though since everything else is going swimmingly with 10.04.3) And I'll certainly be curious to see what the techs say! |
||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Intel man all theway, first test 7 beta13 , version 6.12 on Linux, all valid, and but for the ''MIA'' time, no issues. Self-centered, throttling no issue since it's 24/7 100% crunch on that box. Not contesting that BOINC has to be in total control, simply because one has to be able to boot and have BOINC shutdown processes properly and within reasonable time and know the tasks will resume from last checkpoint, without being dismissed afterwards because of a single restart bit making the quorum match fail.
--//-- |
||
|
Jason1478963
Senior Cruncher United States Joined: Sep 18, 2005 Post Count: 295 Status: Offline Project Badges: |
Kateiacy on the ubuntu upgrade to 11.04 I would make sure you test drive it before you upgrade. I wish I would have checked it out better before I went forward on my laptop. I don't like the unity setup and I seem to have more driver issues on the laptop. It is running an AMD turion on the 64 bit ubuntu 11.04 and had all 3 work units valid on the first batch. I'm still working on 2 for this batch
---------------------------------------- |
||
|
Crystal Pellet
Veteran Cruncher Joined: May 21, 2008 Post Count: 1313 Status: Offline Project Badges: |
That's fascinating, CP! Are those machines running Windows or Linux? What's fascinating about that? They just do the job, for what they are hired Phenom I: Windows Vista 64bit Athlons II: Windows7 64bit |
||
|
Crystal Pellet
Veteran Cruncher Joined: May 21, 2008 Post Count: 1313 Status: Offline Project Badges: |
This is from the only machine with Linux64 running. Intel Dual Core P8400 laptop.
----------------------------------------Because of the amount of inconclusives from other Linux crunchers, it's called now: HOOVER BETA_ BETA20_ ace_ 0000000_ 0083_ 2-- 999763 In Progress 06/08/11 15:23:01 08/08/11 05:47:01 0.00 0.0 / 0.0 |
||
|
wplachy
Senior Cruncher Joined: Sep 4, 2007 Post Count: 423 Status: Offline |
I'm seeing the same results as kateiacy and KerSamson in that my 3 Ubuntu (10.04) PCs have a high rate of invalids (all were Inconclusive) while the Windows PC's have none, as yet. Processing percentage (wall time / CPU time) for all was excelent (99%) running a mix of HCC and DDDT2. I didn't try throttling as all my devices are set for 100%.
----------------------------------------These are the only Invalids the Ubuntu PCs have had in months of running HCC and DDDT. O/S WU's PV Valid Invalid Avg PP Worst PP Best PP Edit: updated counts
Bill P
----------------------------------------[Edit 1 times, last edit by wplachy at Aug 7, 2011 1:52:32 AM] |
||
|
kateiacy
Veteran Cruncher USA Joined: Jan 23, 2010 Post Count: 1027 Status: Offline Project Badges: |
Bill --
----------------------------------------Are your machines Intel or AMD (and if AMD, which class)? Thanks. Kate |
||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
As you typed the same process name twice I'm not 100% clear. You mean the worker process is continuing, but at very low pace and all the time, beyond the few minutes that others have reported? From the 6.12 test, the project folder on linux shows: 1) wcg_beta13_6.12_i686-pc-linux-gnu (the controller) 2) wcg_beta13_graphics_6.12_i686-pc-linux-gnu 3) wcg_beta13_vina_6.12_i686-pc-linux-gnu (the worker) D'oh! Copy/paste failure! Yes, that made it very unclear. (I've edited the original post to make sense.) The first of the 3 controller processes uses small amounts of CPU time no matter what, even when boinc has been told to suspend. Probably not enough to be important, but enough to make it the highest CPU user on a quiet system. |
||
|
wplachy
Senior Cruncher Joined: Sep 4, 2007 Post Count: 423 Status: Offline |
Bill -- Are your machines Intel or AMD (and if AMD, which class)? Thanks. Kate Kate, they are AMD, Athlon's, (1) 3600 & (2) 4800's. Old beasts, but they crank out the work
Bill P
|
||
|
KWSN - A Shrubbery
Master Cruncher Joined: Jan 8, 2006 Post Count: 1585 Status: Offline |
It's (kinda) the AMD that is causing invalids. If a result is sent out on Linux, the wingman will be Linux as well.
----------------------------------------The problem arises because AMD and Intel address their floating point calculations slightly differently and, depending on the project parameters, Intel and AMD processors present different enough results that they don't validate. Once a third unit is sent out, probabilities state it will go to the type with the largest population: Intel. This causes the two Intel results to match up and the AMD result goes invalid. If the third happens to go to an AMD, the Intel result will invalidate. Distributed computing volunteer since September 27, 2000 |
||
|
|