1)
Message boards :
Number crunching :
Interesting credit differences between Linux & Windows wu\'s
(Message 880)
Posted 16 Mar 2006 by tgm Post: It appears that this issue may be somewhat more severe. I'm seeing a similar disparity with the Rosetta production version as well. I need to do some more analysis, but my quick look at results are showing a similar trend. |
2)
Message boards :
Number crunching :
Interesting credit differences between Linux & Windows wu\'s
(Message 879)
Posted 16 Mar 2006 by tgm Post: I'm seeing some very interesting differences with the claimed credit that results from wu's processed on Linux vs. wu's processed under Windows. First, the machines are near identical in terms of computing power. In fact, the cpu_run_time_pref reported by each is identical. The actual CPU seconds processing times and the cpu processing performances I see in my local logs are close enough to each other that they would be considered statistically equal. But for some reason the credit requested by the Windows clients is three (3) times larger then the Linux work units! There are some differences between the Boinc clients (5.2.5 vs. 5.2.13) and the Ralph app version (4.84 vs. 4.92). But is this enough to throw the claimed credit differences that far off? My thought is that someone should look into this and come up with a reasonable explanation. Here are some examples: Windows: WU 15378 WU 15380 WU 15381 Linux: WU 15264 WU 16066 |
3)
Message boards :
Current tests :
Switching between projects with applications removed from memory
(Message 691)
Posted 27 Feb 2006 by tgm Post: I think this is the case when a slower machine (P3/750) takes too long to complete the first model and it gets pre-empted and removed from RAM / VM before even the first checkpoint is reached. I sort of doubt this is the case. I know one of the wu's got up to more than 60% before it crashed. |
4)
Message boards :
Current tests :
Rosetta beta 4.84 wu seems to hang on Linux
(Message 559)
Posted 24 Feb 2006 by tgm Post: I'm seeing some bizare behavior from a wu that is running on a P3 600 Linux (Fedora 4) w/ 768 MB ram. The configuration is set to remove the wu from memory when pre-empted. The work unit gets to a point where it is not doing anything and just sort of sits there and does NOT consume any CPU load. I don't know when it is happening (within the 60 minute cycle), but it is definitely occurring while ralph is running. It's not crashed and boinc isn't crashed either. If I restart the boinc service it restarts where it left off and runs normally. I've been running a Boinc beta (5.2.15) and I just downgraded it back to 5.2.5 to make sure that boinc itself is not the culprit. Hopefully the client does not update anytime soon where I had to shift my configuration on the server to keep the wu in memory in order that Windows wu's can process (another thread). The wu in question is: http://ralph.bakerlab.org/workunit.php?wuid=5558 |
5)
Message boards :
Current tests :
Switching between projects with applications removed from memory
(Message 558)
Posted 24 Feb 2006 by tgm Post: Removing rosetta beta 4.87 work units from memory on one of my windows machines is definitely FAILING with end state client error. This machine is a DUAL PROCESSOR P3 750 w/ 512MB ram running on Windows Server 2003. I have three examples: http://ralph.bakerlab.org/workunit.php?wuid=5559 http://ralph.bakerlab.org/workunit.php?wuid=5560 http://ralph.bakerlab.org/workunit.php?wuid=5561 I have now switched my configuration to keep wu's in memory and performed an update. We'll see what happens. Curiously, I have another wu running on a Fedora box that that is showing some other bizare behavior, but I'll start a new post for this one. |
©2024 University of Washington
http://www.bakerlab.org