Questions and Answers : Getting started : Computation Error
Message board moderation
Author | Message |
---|---|
Send message Joined: 19 Dec 06 Posts: 1 Credit: 402,373 RAC: 0 |
During the last few weeks I have recieved a \'fortran\' computation error that is causing the climate project to stop running. In fact, I have had to start over with a new set of data a couple of times during the past couple of weeks. Below is a copy/paste from the messages section of Bionic. 7/7/2008 11:03:11 AM|climateprediction.net|Reason: Unrecoverable error for result hadsm3fub_jspc_005955331_9 (The device does not recognize the command. (0x16) - exit code 22 (0x16)) 7/7/2008 11:03:17 AM|climateprediction.net|Computation for task hadsm3fub_jspc_005955331_9 finished 7/7/2008 11:03:17 AM|climateprediction.net|Output file hadsm3fub_jspc_005955331_9_1.zip for task hadsm3fub_jspc_005955331_9 absent 7/7/2008 11:03:17 AM|climateprediction.net|Output file hadsm3fub_jspc_005955331_9_2.zip for task hadsm3fub_jspc_005955331_9 absent 7/7/2008 11:03:17 AM|climateprediction.net|Output file hadsm3fub_jspc_005955331_9_3.zip for task hadsm3fub_jspc_005955331_9 absent Any insight? Thanks, Dennis |
Send message Joined: 5 Sep 04 Posts: 7629 Credit: 24,240,330 RAC: 0 |
Hi Dennis The error 22 seems to be a general one, which gets applied to a lot of failure modes. The last 3 lines just mean that BOINC was told at the start to expect those files to exist when the model completed. The model didn\'t, so the files haven\'t been created. As for Fortran errors, this is still a bit of a mystery, as it only happens to a few people, and then not all of the time. It\'s possible that the program couldn\'t access the hard disk when it needed it. As this program originates from a supercomputer program, it was never envisaged that it would get interrupted by lots of other programs, as happens on a desktop computer. So, when it can\'t get access to it\'s HD files, it doesn\'t wait around. It just crashes. It has been mentioned quite a few times on these boards, so an advanced search for fortran (at the top of the page), may give you some insights. Best I can tell you is: ALWAYS exit from BOINC before shutting down the computer. ALWAYS Suspend BOINC before running other resource intensive program. (Video streaming/editing, games, etc.) Have a look at the README files linked from my sig below. Backups: Here |
Send message Joined: 15 May 09 Posts: 4541 Credit: 19,039,635 RAC: 18,944 |
Every time I try to run a unit using climate prediction.net I get the computation error message after only a minute or so. I do not have any other resource intensive programs running. I am using Mandriva2009.1 It is the same computer that I have run the climate-prediction.net software on before. I am able to run other BOINC client tasks without issue.Where do I look to try and find an answer as to how to resolve this problem. It has been persistent for several months. Dave |
Send message Joined: 5 Aug 04 Posts: 1283 Credit: 15,824,334 RAC: 0 |
Hi Dave, The problem is that your computer has only been receiving HADAM3P tasks. They require SSE2 but your Athlon doesn't have that capability. The project team are working on finding a way to ensure those tasks are only sent to systems which can run them. In the meantime the only solution is for you to alter your project preferences to exclude that model type. I'd suggest that you also exclude HADAM3 because of its high memory requirement. "The ultimate test of a moral society is the kind of world that it leaves to its children." - Dietrich Bonhoeffer |
©2024 cpdn.org