41)
Message boards :
Number crunching :
Problems with Rosetta version 5.93
(Message 50611)
Posted 12 Jan 2008 by Astro Post: It's been three days without any new watchdog errors. Here's my scoreboard for 5.93. Personally, I wonder what's different between my hosts and those of users like Luuklag who also has an AMD64 host but IS getting computation errors. I haven't seen one computation error yet, so something must be different. |
42)
Message boards :
Number crunching :
Problems with Rosetta version 5.93
(Message 50592)
Posted 11 Jan 2008 by Astro Post: The entire computer has crashed and switched off whenever the Rosetta 5.93 Beta has been running. Michael, If you puter is shutting down, then some basic computer parameter is being triggered. Have you checked for dust build up on your CPU heatsink, RAM, Power Supply Unit, etc lately? I believe what you are experiencing is heat related, or atleast it's my "best guess". I can't think of anything in the Seti or Rosetta (or any other Boinc Science app) that would cause a system shutdown. When any application runs it creates heat. depending upon the application, some apps work your processor harder than others, but all try to get 100% use out of the processor. The more efficient an app becomes, the hotter the processor should get as it's doing more in less time. anyway, I'd check the temps and/or dust accumulation inside your puter. tony |
43)
Message boards :
Number crunching :
Problems with Rosetta version 5.93
(Message 50528)
Posted 10 Jan 2008 by Astro Post: My AMD64 X2 4800 using Windows had TWO "1zpy" wus ended by the watchdog yesterday. At this point, I don't see the need to link to the results as they must have PLENTY of samples to work with. |
44)
Message boards :
Number crunching :
"Host" stat on RAH homepage...what about Cores?
(Message 50513)
Posted 9 Jan 2008 by Astro Post: Almost sounds like your asking for opinions. That said, there's boatloads of stats available from Boincstats etc. I also think that something in the "Boinc" back end would have to be changed to support this. It would need to be smart enough to decide if a "p4 2.8" is H/T capable, or basically a full definition of available cpu architectures. would cpuid string provide enough info?? In the end, stats are interesting, some are more important than others to some people, and the importance changes sometimes daily. So, Would it be nice???? Well, any new stats would be nice. Is it/should it be a high priority for the Boinc team??? probably not. I know that Dr. A and Rom are giving much of the coding to volunteers for certain features (like the forums). anyway. It'd be nice....yes. 2c inserted |
45)
Message boards :
Number crunching :
Problems with Rosetta version 5.93
(Message 50487)
Posted 9 Jan 2008 by Astro Post: My 4800 using linux had a watchdog ended WU, but it too was valid and credited. Here's the current scoreboard of 5.93 for my hosts. It seems consistent that it's the 1zpy wus that I have issues with. Perhaps, since I get credit and the wu is considered valid, that this is more of an "informational message", rather than an "error"??? |
46)
Message boards :
Number crunching :
Anyone know how the Phenom 9600 is as a cruncher?
(Message 50486)
Posted 9 Jan 2008 by Astro Post: http://www.lagoom.com/AMD_Phenom_9600_Quad-Core_2.3GHz/HD960ZWCGDBOX/partinfo-id-579600.html Black Edition $247.50. Says "instock" |
47)
Message boards :
Number crunching :
Problems with Rosetta version 5.93
(Message 50478)
Posted 9 Jan 2008 by Astro Post: Yup, Looks like the 1zpy's are getting their normal "watchdog" endings (but in this case valid and credited. resultid=131958175 (this is the only one for me so far. I.E no other errors of any kind except this one, so not so bad) |
48)
Message boards :
Number crunching :
Newbie Q&A, if you're new, have a view!
(Message 50456)
Posted 8 Jan 2008 by Astro Post: Is there anything special I need to do to report my finished work? Nope. The reporting is done per your Connection Interval. It hopes to report more than one at a time to reduce loads on the server. However, If you go to the "projects" page, highlight "rosetta@home" (in the right hand box), then click "update" (in the left hand box), you can manually report them/it. This is called a "project update". But no, there's nothing special you "need" to do. hope this helps tony |
49)
Message boards :
Number crunching :
Problems with Rosetta version 5.93
(Message 50448)
Posted 8 Jan 2008 by Astro Post: Perhaps if this thread were called "NO Problems with Rosetta version 5.93" then there'd be more posts (knock on wood). I'm not seeing many reports of errors. Looking at my own hosts, I've not had one error either with Windows or Linux. Wonder what would happen if they fed us some 1zpy boinc twist rings wus???. While this updated chart only shows a small number of 5.93's I'd have thought I'd have seen some sign of trouble by now, expecially if you consider the numbers for 5.90 (in windows) and 5.91 (in Linux). I've highlighted 5.93 in red. (my puters are dual boot using both windows and linux as designated by the "l" or "w" following the number. The number is the AMD designation, so put AMD64 in front of the 2800 and 3700, and put AMD64 X2 in front of the 4800, 5200, and 6000.) |
50)
Message boards :
Number crunching :
Anyone else getting computation errors?
(Message 50314)
Posted 4 Jan 2008 by Astro Post: i thought memtest86 was a free download? Of course it is free http://www.memtest.org/ Perhaps they can't burn a CD or floppy and have to have it done for them??? |
51)
Message boards :
Number crunching :
Problems with version 5.90/5.91
(Message 50208)
Posted 31 Dec 2007 by Astro Post: resultid=128931062 left me with a windows crash box prompt asking if I wanted to report it to microshaft. |
52)
Message boards :
Cafe Rosetta :
Personal Milestones
(Message 50185)
Posted 30 Dec 2007 by Astro Post: 300K Rosie, 1/2 way to final goal |
53)
Message boards :
Number crunching :
Better Graphics and resolution in Rosetta?
(Message 50162)
Posted 29 Dec 2007 by Astro Post: Quick newbie question: do the graphics influence the computation, that is, the results? Only in that it needs memory to run, and that might slow down your computations of the real work by some small percentage (depending on video card/integrated graphics, amount of ram, age of machine, etc etc). Some users don't use the screensaver or look at the graphic to increase the amount of work done, but I don't think it's to any great effect. |
54)
Message boards :
Number crunching :
Better Graphics and resolution in Rosetta?
(Message 50139)
Posted 28 Dec 2007 by Astro Post: I love the way the chains in Rosetta look, but I would like them to display in higher resolution! If you click on "participants" at the top of this page, it'll take you to "your account". click the link "rosetta@home preferences", then you can play with the graphics settings there to see if you can make it appear satisfactorially to you. The settings will update on your machine from those "web based" settings after it calls home for new work, to report work, or if you do a "project update" (go to "projects" tab on your boinc manager, highlight "rosetta" in the right hand box, then click "update" in the left column). I don't know of a "hi/lo res" setting. good luck tony |
55)
Message boards :
Number crunching :
Problems with version 5.90/5.91
(Message 50134)
Posted 28 Dec 2007 by Astro Post: Here's a list of ALL the watchdog ended tasks for all hosts/OSes using 5.90 or 5.91 since Nov 26th 2007... Do they have something in common??? 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_12692_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_12604_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_12581_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_12309_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_12265_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_12266_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_12225_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_12206_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_11882_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_10396_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_10466_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_14658_0 1zpy__BOINC_TWIST_RINGS_TWIST_ANGLE_SYMM_FOLD_AND_DOCK_RELAX-1zpy_-native__2477_115770_0 1zpy__BOINC_TWIST_RINGS_TWIST_ANGLE_SYMM_FOLD_AND_DOCK_RELAX-1zpy_-native__2477_115774_0 1zpy__BOINC_TWIST_RINGS_TWIST_ANGLE_SYMM_FOLD_AND_DOCK_RELAX-1zpy_-native__2477_121452_0 1zpy__BOINC_TWIST_RINGS_TWIST_ANGLE_SYMM_FOLD_AND_DOCK_RELAX-1zpy_-native__2477_156769_0 1zpy__BOINC_TWIST_RINGS_TWIST_ANGLE_SYMM_FOLD_AND_DOCK_RELAX-1zpy_-native__2477_161145_0 1zpy__BOINC_TWIST_RINGS_TWIST_ANGLE_SYMM_FOLD_AND_DOCK_RELAX-1zpy_-native__2477_183748_0 1zpy__BOINC_TWIST_RINGS_TWIST_ANGLE_SYMM_FOLD_AND_DOCK_RELAX-1zpy_-native__2477_36172_0 1zpy__BOINC_TWIST_RINGS_TWIST_ANGLE_SYMM_FOLD_AND_DOCK_RELAX-1zpy_-native__2477_148361_0 1zpy__BOINC_TWIST_RINGS_TWIST_ANGLE_SYMM_FOLD_AND_DOCK_RELAX-1zpy_-native__2477_144472_0 1zpy__BOINC_TWIST_RINGS_TWIST_ANGLE_SYMM_FOLD_AND_DOCK_RELAX-1zpy_-native__2477_144384_0 1zpy__BOINC_TWIST_RINGS_TWIST_ANGLE_SYMM_FOLD_AND_DOCK_RELAX-1zpy_-native__2477_143544_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_11173_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_11055_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_11186_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_11059_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_12138_0 1zpy__BOINC_TWIST_RINGS_TWIST_ANGLE_SYMM_FOLD_AND_DOCK_RELAX-1zpy_-native__2477_15985_1 1zpy__BOINC_TWIST_RINGS_TWIST_ANGLE_SYMM_FOLD_AND_DOCK_RELAX-1zpy_-native__2477_74554_0 1zpy__BOINC_TWIST_RINGS_TWIST_ANGLE_SYMM_FOLD_AND_DOCK_RELAX-1zpy_-native__2477_74522_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_10255_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_10233_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_10229_0 1zpy__BOINC_TWIST_RINGS_TWIST_ANGLE_SYMM_FOLD_AND_DOCK_RELAX-1zpy_-native__2477_73415_0 1zpy__BOINC_TWIST_RINGS_TWIST_ANGLE_SYMM_FOLD_AND_DOCK_RELAX-1zpy_-native__2477_73402_0 1zpy__BOINC_TWIST_RINGS_TWIST_ANGLE_SYMM_FOLD_AND_DOCK_RELAX-1zpy_-native__2477_73393_0 1zpy__BOINC_TWIST_RINGS_TWIST_ANGLE_SYMM_FOLD_AND_DOCK_RELAX-1zpy_-native__2477_50756_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_13658_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_13656_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_13652_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_11899_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_11965_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_11959_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_11976_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_10875_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_10838_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_10826_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_10090_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_11281_0 1zpy__BOINC_TWIST_RINGS_TWIST_ANGLE_SYMM_FOLD_AND_DOCK_RELAX-1zpy_-native__2477_30600_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_12760_0 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2470_12740_0 There are NO watch dog ended tasks for apps earlier than 5.90/5.91 |
56)
Message boards :
Number crunching :
Problems with version 5.90/5.91
(Message 50133)
Posted 28 Dec 2007 by Astro Post: wooohoooo I got the 100th post. OK OK back to bidness. There's definitely something wrong with 5.90/5.91. Either the watchdog is set to a value below what it should be causing the watchdog to end a task, or the app itself has an error, or the Job type has an error, or it might be something else entirely. How can I make this claim??? your ask?? Below is a chart of all my recorded work from app 5.82 upwards (except 5.90 for linux which is "known bad"). You can see that the former apps didn't have near the incidence of watchdog time out that 5.91 does. Note how there aren't any watchdog errors on earlier versions and almost no compute errors for earlier windows versions. The "compute" errors for earlier linux versions are tied to the "loss of internet" errors which I reported earlier. For host names 6000l is AMD64 X2 6000 under linux, 6000w is the same under windows, etc etc etc for the others. |
57)
Message boards :
Number crunching :
Anyone else getting computation errors?
(Message 50089)
Posted 26 Dec 2007 by Astro Post: I also have recently received multiple "computation error" results which also brought up the WinXP "Report this problem" window when a program crashes. Happened on two different machines on two different networks. I had the exact same thing with resultid=128948581, though it's the only one I've had in months. It did happen today like yours. |
58)
Message boards :
Number crunching :
Problems with version 5.90/5.91
(Message 50016)
Posted 24 Dec 2007 by Astro Post: There's many kinds of that similar wu. These new ones are "twist rings, twist angle, symm fold and dock relax .... 2477 THe emboldened parts are the NEW parts. THe older ones were 2470 as well, and those ran well on my lower mem systems. 1zpy__BOINC_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2474_?_0 1zpy__BOINC_TWIST_RINGS_MORE_SLIDESYMM_FOLD_AND_DOCK-1zpy_-native__2476_?_0 1zpy__BOINC_TWIST_ANGLE_SYMM_FOLD_AND_DOCK-1zpy_-native__2474_?_0 1zpy__BOINC_TWIST_ANGLE_MORE_SLIDESYMM_FOLD_AND_DOCK-1zpy_-native__2476_?_0 1zpy__BOINC_LESSCYCLES_TWIST_RINGS_TWIST_ANGLE_SYMM_FOLD_AND_DOCK-1zpy_-native__2471_?_0 1zpy__BOINC_LESSCYCLES_TWIST_RINGS_SYMM_FOLD_AND_DOCK-1zpy_-native__2471_?_0 1g2z__BOINC_TWIST_RINGS_TWIST_ANGLE_SYMM_FOLD_AND_DOCK_RELAX-1g2z_-crystal_foldanddock__2469_?_0 Note: the last one is different from the new one because of the "crystal_foldand dock ????" and 1g2z |
59)
Message boards :
Number crunching :
Problems with version 5.90/5.91
(Message 50010)
Posted 24 Dec 2007 by Astro Post: I have a single core CPU, and I have just started getting those limited memory errors when attempting to request new work from Rosetta. Jack, You have 768 MB of "installed" memory. This system requirement refers to "useable" memory (memory not reserved/taken by other things like the operating system). My best guess is that the new wus Izpy Boinc Twist Rings Twist Angles Symm Fold and Dock Relax 1zpy 2477 ? are requiring more memory than you/I/MOST(by most, I mean I think the vast majority of hosts connected to rosetta have 512K or less, although that percentage will lessen over time) of the "normal" crunchers out there, and they don't have any other work queued up for us to participate on. So in effect they're saying "we don't need you at the moment, please come back when we do". One way to stay productive would be to have a cache as large as humanly (computerly???) possible to carry us through these times. You/everyone affected could keep a close eye on the fora and the first time this is mentioned then race to increase your "cpu run time pref" so you could also maintain work for the period in which they have none to offer. Atleast that's my guess. |
60)
Message boards :
Number crunching :
System requirements????
(Message 49996)
Posted 24 Dec 2007 by Astro Post: Here's what I'm trying. Set cpu run time to 24 hours. Update rosetta on that machine. Confirm successful contact with server. Set host to "network activity suspended". Change run time pref back on the website. Hopefully, they'll find some "normal" (whatever that is) size wus for us to use. |
©2024 University of Washington
https://www.bakerlab.org