Questions and Answers :
Unix/Linux :
Climateprediction/BBC model not starting
Message board moderation
Author | Message |
---|---|
Send message Joined: 10 Jun 06 Posts: 1 Credit: 653 RAC: 0 |
Hi all, I have recently finished my previous workuint (it took quite a long time) and I downloaded a new one ... I was running an old version of BOINC (6-7 months ago) and I realized that the new workuint did not start processing. I did update to the latest BOINC version (5.4.9) but I does not work either ... after several minutes it reports an error and downloads a new one ... I have tried a fresh installation and it is always the same issue ... download, try for a few minutes and report an error ... I am running Linux Slackware 10.1 (Kernel 2.6.11.8) ... glibc 2.3.2 ... Any help would be appreciated. Thanks in advance. Cheers, J. |
Send message Joined: 5 Aug 04 Posts: 1496 Credit: 95,522,203 RAC: 0 |
Hi, jako2, welcome to the Boards. Don\'t know where credit went for your previous work; your account shows none. Did you re-register when you upgraded to 5.4.9? (If so, that would account for it.) This is the likely reason for current failures: AMD Athlon(tm) XP 1600+ Memory 255.48 MB The CPU is marginal and memory is half the recommended amount to run the Coupled Model. Edit: As you are running Linux, you *might* get by with your rig -- but a Coupled Model would probably take the better part of a year, running 24/7. Also, the hard drive would take a beating, especially if you did other work on the machine. "We have met the enemy and he is us." -- Pogo Greetings from coastal Washington state, the scenic US Pacific Northwest. |
Send message Joined: 21 Nov 05 Posts: 2 Credit: 231,636 RAC: 0 |
I am sorry for the mess ... this one (\"jako\") is the original profile ... the other one is a new box running Windows which I registered a few weeks ago. I add this comment, should the hardware configuration has something to do with the problem ... Thanks in advance. |
Send message Joined: 5 Aug 04 Posts: 1496 Credit: 95,522,203 RAC: 0 |
I am sorry for the mess ... this one (\"jako\") is the original profile ... the other one is a new box running Windows which I registered a few weeks ago. Hi, jako, Thanks for the clarification. Yes, the hardware configuration has everything to do with running the Model. Your P4 is short of memory. The Pentium M is powerful enough but please be sure to keep it elevated above the desk to allow air to flow under it for cooling. (Some laptops are notorious for overheating when running 100% for extended periods.) It\'s also important to keep the vents and inside cleared of accumulated dust. Also, make regular backups of the entire boinc folder, at least every week, so you can recover after crashes -- and, I fear, you will experience crashes. In any event, these Models cover 160 years of both atmosphere and ocean, from very high to very deep. It takes a long time to process all that data. For example, Runs on my A64 X2 4400+ machines require nearly 100 days, running all the time, to finish. A P4 was recently rebuilt and made into Pentium D 940; it\'s a bit slower than the AMDs and will take a few days longer to finish its Runs... (Those examples are simply to indicate what is required for these very demanding Models.) The continuing errors you\'ve experienced are trying to tell you something, eh? The Sulphur Cycle part of the Project is completed. It\'s purpose was to develop input-data for the Coupled Models. (Same with the 200-year ocean Spinups some of us ran.) The one you have in progress is a long way from completion. I\'m very uncomfortable recommending that someone not try to run CPDN Models. It could be misconstrued as rejection -- and it certainly isn\'t that. However, you might get more satisfaction running one of the other distributed science Projects on your slowest machines. You\'ll have your hands full trying to finish a Model on your laptop. (Many people use laptops in the Project, but we don\'t recommend it because of the potential problems. It\'s up to you.) Best of luck, jako. "We have met the enemy and he is us." -- Pogo Greetings from coastal Washington state, the scenic US Pacific Northwest. |
Send message Joined: 21 Nov 05 Posts: 2 Credit: 231,636 RAC: 0 |
Hi AstroWX, thanks for the prompt answer ... so, it seems that none of my computers are powerful enough to run the new model ... it is a pity, for this is the ONE project I wanted to contribute to ... I had already decided to increase my P4\'s memory (actually, when trying to start the model I did not see any memory problem in the box) ... but it seems that it could not be enough, though. I will let my laptop finish the currently running workunit (sulphur cycle 4.22, 46% completed) but I won\'t request a new one ... Thanks for the clarification, anyway. Cheers, Jacobo. I am sorry for the mess ... this one (\"jako\") is the original profile ... the other one is a new box running Windows which I registered a few weeks ago. |
©2024 cpdn.org