 |
|
Questions and Answers :
Windows :
Visual Fortran runtime-error
Message board moderation
| Author | Message |
|---|---|
|
Send message Joined: 6 Dec 05 Posts: 7 Credit: 671,245 RAC: 0 |
I had been getting this error for several weeks and due to other OS probs I did a complete wipe and OS reinstall. Been up and running 1 day and still get the error:  Can anyone tell me what causes this and if there is a fix? I also have lost several work units due to computational errors: 6/30/2007 5:25:25 AM|climateprediction.net|Reason: Unrecoverable error for result hadsm3fub_0060_005892471_4 (The device does not recognize the command. (0x16) - exit code 22 (0x16)) Anyone know why this occurs and if it may be related to my PC? Running BOINC V:5.8.16 Win XP Pro SP2, AMD64 x2 6000+, 2 gig PC6400, nVidia 8600 GTS. |
|
Send message Joined: 5 Sep 04 Posts: 7629 Credit: 24,240,330 RAC: 0 |
The Fortran error is most likely unrelated to the error 22. The latter has \'probably\' been found, but the \"fixed\" version is still in beta testing. The image shown mentions a write error, so there are at least 2 possibilities: 1) Your hd is failing 2) You\'re running something else at the same time that is using the hd a lot, and causing a conflict with the model when it trys to write to the hd. (Video/audio streaming/editing, for instance, but there are lots of others.) |
|
Send message Joined: 6 Dec 05 Posts: 7 Credit: 671,245 RAC: 0 |
Thanks for the reply. As far as HDD usage while BOINC runs I do assure that it has all resources available to it. The only thing that does run is a couple internet apps such as weather, an IM client, that is seldom used, and a CPU temp monitor. Thing about the HDD is that it is relatively new and I had just used DBaN for a DOD wipe with no reported HDD errors as it would normally do. This error has only began since installation of the latest BOINC client. I dunno? Will look further into the health of my HDD.
|
 astroWXSend message Joined: 5 Aug 04 Posts: 1496 Credit: 95,522,203 RAC: 0 |
Is the machine overclocked? If so, does it pass Memtest-86, error-free, for several hours? Prime-95 Torture Test, one copy per core, for 24 hours? "We have met the enemy and he is us." -- Pogo Greetings from coastal Washington state, the scenic US Pacific Northwest. |
|
Send message Joined: 6 Dec 05 Posts: 7 Credit: 671,245 RAC: 0 |
Not overclocked! Running everything at stock speeds. I\'m going to run memtest and HDDscan later on. |
|
Send message Joined: 6 Dec 05 Posts: 7 Credit: 671,245 RAC: 0 |
HDD checked out fine. Looks like I have a bad stick of RAM! Memtest showed something like 38000 errors and counting when I exited. New Ram only 2 months in use! Does Visual Fortran write to RAM? If it does then that could very well be the problem. Thanks! |
|
Send message Joined: 5 Sep 04 Posts: 7629 Credit: 24,240,330 RAC: 0 |
EEEuuuuw. Nasty. Sorry to hear about that. The problem may well be the program itself, as it\'s written in Fortran. All one million plus lines of it. |
mo.v Send message Joined: 29 Sep 04 Posts: 2363 Credit: 14,611,758 RAC: 0 |
Assuming that you have 2 x 1Gb sticks of RAM, you could remove the faulty one, run the model/s on the remaining RAM and see whether the Fortran error messages disappear. On XP, 1Gb is enough to run two models side-by-side. Cpdn news |
|
Send message Joined: 5 Jan 06 Posts: 1 Credit: 28,241 RAC: 0 |
I have the same error since a few weeks. It is kind of nasty, because a click on OK gives me the same error again and again. I have to stop BOINC at all until reboot to get rid of it. |
MikeMarsUK Send message Joined: 13 Jan 06 Posts: 1498 Credit: 15,613,038 RAC: 0 |
Have you tried Prime95\'s torture test and memcheck to make sure the PC hardware is happy? I'm a volunteer and my views are my own. News and Announcements and FAQ |
|
Send message Joined: 6 Dec 05 Posts: 7 Credit: 671,245 RAC: 0 |
I have found that when the error occurs a reboot allows BOINC to run for 8+ hours error free. I\'m almost convinced that faulty RAM is the culprit as a reboot would flush the RAM. |
|
Send message Joined: 6 Dec 05 Posts: 7 Credit: 671,245 RAC: 0 |
Assuming that you have 2 x 1Gb sticks of RAM, you could remove the faulty one, run the model/s on the remaining RAM and see whether the Fortran error messages disappear. On XP, 1Gb is enough to run two models side-by-side. OK, gave that a try and still received the error! I\'ll get new RAM, in about 2 weeks, and wait untill then to actually find out if it is a RAM issue. Thanks for the suggestion. |
|
Send message Joined: 17 Jul 07 Posts: 6 Credit: 49,684 RAC: 0 |
I got a fortran error too. erm, error during write C;/...path to relevant files. I had assumed that it was a programming error specific to the climatepredition task as thats all the rest of the error mentions. Im new to this, so if it happens again what should I be looking out for? |
|
MikeMarsUK Send message Joined: 13 Jan 06 Posts: 1498 Credit: 15,613,038 RAC: 0 |
Hi there, If you follow the link to your crashed result, you\'ll see a page full of error log. Most of it is irrelevant (minor warning messages), but at the end is this section: CPDN process is not running, exiting, bRetVal = 1, checkPID=0, selfPID=2536, iMonCtr=1 The \'access is denied\' bit is odd. Do you recall what was happening on the PC at the time the model crashed, for example something running and taking 100% of CPU time (games, video encoding, etc), or perhaps running an update of some kind? If you follow the link in my signature, you\'ll find several several posts which contain ideas about preventing crashes and doing backups. I'm a volunteer and my views are my own. News and Announcements and FAQ |
|
Send message Joined: 17 Jul 07 Posts: 6 Credit: 49,684 RAC: 0 |
Thanks. Hmm I was running some stuff that canes the processor a bit. For example a programme that reads text back to me. I dont have anything that should fight with a fortran programme though, as far as I know... |
|
MikeMarsUK Send message Joined: 13 Jan 06 Posts: 1498 Credit: 15,613,038 RAC: 0 |
Anything which uses 100% of the CPU for more than 15 minutes is potentially a problem, I wouldn\'t have thought a dictation program would do that though ... I'm a volunteer and my views are my own. News and Announcements and FAQ |
|
Send message Joined: 6 Dec 05 Posts: 7 Credit: 671,245 RAC: 0 |
Assuming that you have 2 x 1Gb sticks of RAM, you could remove the faulty one, run the model/s on the remaining RAM and see whether the Fortran error messages disappear. On XP, 1Gb is enough to run two models side-by-side. Want to let you know that some new, error free, RAM seems to have solved the Fortran error problem. BOINC has been running error free for at least 72 hours now. Used to get only 9 error free hours max. I\'m gonna call this problem resolved! Thanks for the input. |
|
MikeMarsUK Send message Joined: 13 Jan 06 Posts: 1498 Credit: 15,613,038 RAC: 0 |
|
batan Send message Joined: 25 Aug 07 Posts: 17 Credit: 80,535 RAC: 0 |
I just encountered this same Fortran error (-38) with a slab model. 31.08.2007 07:26:16|climateprediction.net|Sending scheduler request: To send trickle-up message 31.08.2007 07:26:16|climateprediction.net|(not requesting new work or reporting completed tasks) 31.08.2007 07:26:21|climateprediction.net|Scheduler RPC succeeded [server version 509] 31.08.2007 07:26:31|climateprediction.net|Sending scheduler request: To send trickle-up message 31.08.2007 07:26:31|climateprediction.net|(not requesting new work or reporting completed tasks) 31.08.2007 07:26:36|climateprediction.net|Scheduler RPC succeeded [server version 509] 31.08.2007 07:26:46|climateprediction.net|Sending scheduler request: To send trickle-up message 31.08.2007 07:26:46|climateprediction.net|(not requesting new work or reporting completed tasks) 31.08.2007 07:26:51|climateprediction.net|Scheduler RPC succeeded [server version 509] 31.08.2007 07:28:25|climateprediction.net|Deferring communication for 1 min 0 sec 31.08.2007 07:28:25|climateprediction.net|Reason: Unrecoverable error for result hadsm3fub_0278_005900648_5 (Das Gerät erkennt den Befehl nicht. (0x16) - exit code 22 (0x16)) 31.08.2007 07:28:31|climateprediction.net|Computation for task hadsm3fub_0278_005900648_5 finished 31.08.2007 07:28:31|climateprediction.net|Output file hadsm3fub_0278_005900648_5_1.zip for task hadsm3fub_0278_005900648_5 absent 31.08.2007 07:28:31|climateprediction.net|Output file hadsm3fub_0278_005900648_5_2.zip for task hadsm3fub_0278_005900648_5 absent 31.08.2007 07:28:31|climateprediction.net|Output file hadsm3fub_0278_005900648_5_3.zip for task hadsm3fub_0278_005900648_5 absent This happened immediately after logging on again to windows in the morning (no screensaver, but logon screen appears after absence). According to the last trickle the error might well have occured during the night. But actually I think it occured as the next trickle was due to be sent (maybe the first phase just completed - I expected it to complete this morning, but a bit later). I clicked OK three times on that Fortran error dialogue after which it disappeared. BOINC was still running. I checked the graph (world map) and it was all blue, but the window closed by itself after seconds, so I didn\'t have time to check if phase 1 was complete. The harddisks are only 7 months old and I have never had disk problems on this computer. I am also not aware of any RAM errors. The PC is perfectly stable, well cooled and not overclocked. On the other core there is a hadam3 model running. I also had einstein running with a lower priority (which now took over the unused core). I also have BURP in the project list (which is beta), but it had no work, so it wasn\'t running. I have an anti-virus program running, but it didn\'t report anything. Now I excluded the BOINC directory from its configuration anyway. I then updated the project to inform the server, after which the model disappeared from the task list. Apparently no model data were submitted. In the BOINC folder the model data seems to be still there - so should I mail the contents of the folder (zipped) to some mail address? Or are those files uninteresting, since the model hasn\'t even completed phase 1? |
|
batan Send message Joined: 25 Aug 07 Posts: 17 Credit: 80,535 RAC: 0 |
Since others seem to have had faulty RAM with this error, I\'ve run MemTest (with and without BOINC running) but so far memory seems to be ok. They recommend running the test application for several hours though, so I\'ll continue testing. |
©2024 climateprediction.net
