Questions and Answers :
Windows :
Visual Fortran run-time error
Message board moderation
Author | Message |
---|---|
Send message Joined: 14 Nov 06 Posts: 1 Credit: 158,270 RAC: 0 |
Hi,I got the last time the next error: "forrtl: severe (38): error during write, unit 8, file C:ProgramDataBoincProjectsclimateprediction.nethadam3p_mjq7_1966_2_006057761tmppipe_dummy. Image PC Routine Line Source hadam3p_um_6.06_w 0084E86A Unknown Unknown Unknown Stack trace terminated abnormally. Status: miscalculation" So that is the problem, what can I do now? CPDC download then a new version and I get the same problem. Asclepios. |
Send message Joined: 5 Aug 04 Posts: 1283 Credit: 15,824,334 RAC: 0 |
The most likely cause is your anti-virus program. The Fortran compiler is very unforgiving when it tries to write to a locked file, and AV programs usually set an exclusive lock when they scan a file. Setting up an AV scan exclusion on C:\ProgramData\Boinc\Projects usually sorts it out. Note that some AVs (e.g. Norton) have separate manual and automatic exclusion lists. "The ultimate test of a moral society is the kind of world that it leaves to its children." - Dietrich Bonhoeffer |
Send message Joined: 24 Sep 04 Posts: 3 Credit: 226,837 RAC: 0 |
Hi, I too am getting Intel(r) Visual Fortran run-time errors: Forrtl: severe (24) end-of-file during read, unit 9. This has happened several times (with completion %age in single figures. Each time I have reset the project, this time it has reached nearly 26% but has done it again. I see from a previous answer that it is suggested that it may be due to anti-virus programme. The error occurred first when I was using Norton 360 but it is now happening when I am using McAfee. Any suggestions? |
Send message Joined: 5 Sep 04 Posts: 7629 Credit: 24,240,330 RAC: 0 |
As has been said many times, this is an old problem that only affects some computers some of the time. There\'s still no definitive answer; just make sure that nothing is stopping the program from reading or writing to the hard disk when ever it wants to. This is a massive Fortran program designed and written for supercomputers, and it doesn\'t like being interrupted at certain critical times, such as when it wants the main storage device. Also, there\'s no point in resetting BOINC each time this happens; it\'s not the program, it\'s your computer/computer usage. Resetting just wastes the model, as well as time downloading all of the programs and files again. Make regular backups, and restore from the latest after a failure. Backups: Here |
Send message Joined: 15 Jun 05 Posts: 1 Credit: 8,795,070 RAC: 0 |
I\'ve also recently run into this problem (specifically, a forrtl error not being able to write to a pipe_dummy file in the climateprediction.net data dir) and seem to have resolved it (so far...). Basically, even after quitting BOINC manager, there were still a couple hadxxx processes still running (as seen in Task Manager on WinXP). No idea how they were left behind but killing them allowed me to do a rename on the pipe_dummy file, something I could not do before that (I\'d get a \"file in use\" message). There was also what looked like an orphan BOINC Tray Manager (I think) process which I also killed, but my money is on the hadxxx processes. (Note that \"hadxxx\" is supposed to be shorthand for something like hadsm3_6.07_windows_intelx86.exe.) After undoing the rename and restarting BOINC, the task resumed w/o any runtime errors. This is certainly not the answer to all the variations of this problem that I\'ve seen here, but I haven\'t seen this answer so I figured I\'d share. Cheers, Dave |
©2024 climateprediction.net