I don\'t know what this statment means...please give me a clue
TIA
Alan
27/08/2007 08:26:12||Starting BOINC client version 5.8.16 for windows_intelx86
27/08/2007 08:26:12||log flags: task, file_xfer, sched_ops
27/08/2007 08:26:12||Libraries: libcurl/7.16.0 OpenSSL/0.9.8a zlib/1.2.3
27/08/2007 08:26:12||Data directory: C:\\Program Files\\BOINC
27/08/2007 08:26:12||Processor: 2 GenuineIntel Intel(R) Pentium(R) 4 CPU 3.00GHz [x86 Family 15 Model 2 Stepping 9] [fpu tsc sse sse2 mmx]
27/08/2007 08:26:12||Memory: 1022.98 MB physical, 2.40 GB virtual
27/08/2007 08:26:12||Disk: 232.82 GB total, 222.07 GB free
27/08/2007 08:26:12|climateprediction.net|URL: http://climateprediction.net/; Computer ID: 530152; location: home; project prefs: default
27/08/2007 08:26:12||General prefs: from http://bbc.cpdn.org/ (last modified 2006-07-06 02:12:54)
27/08/2007 08:26:12||Host location: none
27/08/2007 08:26:12||General prefs: using your defaults
27/08/2007 08:26:12||Reading preferences override file
27/08/2007 08:26:26|climateprediction.net|Restarting task hadcm3inct_cll3_1920_160_75868179_2 using hadcm3i version 544
28/08/2007 00:35:55|climateprediction.net|Deferring communication for 1 min 0 sec
28/08/2007 00:35:55|climateprediction.net|Reason: Unrecoverable error for result hadcm3inct_cll3_1920_160_75868179_2 (The device does not recognize the command. (0x16) - exit code 22 (0x16))
28/08/2007 00:35:55|climateprediction.net|Computation for task hadcm3inct_cll3_1920_160_75868179_2 finished
28/08/2007 00:35:55|climateprediction.net|Output file hadcm3inct_cll3_1920_160_75868179_2_4.zip for task hadcm3inct_cll3_1920_160_75868179_2 absent
28/08/2007 00:35:55|climateprediction.net|Output file hadcm3inct_cll3_1920_160_75868179_2_5.zip for task hadcm3inct_cll3_1920_160_75868179_2 absent
28/08/2007 00:35:55|climateprediction.net|Output file hadcm3inct_cll3_1920_160_75868179_2_6.zip for task hadcm3inct_cll3_1920_160_75868179_2 absent
28/08/2007 00:35:55|climateprediction.net|Output file hadcm3inct_cll3_1920_160_75868179_2_7.zip for task hadcm3inct_cll3_1920_160_75868179_2 absent
28/08/2007 00:35:55|climateprediction.net|Output file hadcm3inct_cll3_1920_160_75868179_2_8.zip for task hadcm3inct_cll3_1920_160_75868179_2 absent
28/08/2007 00:35:55|climateprediction.net|Output file hadcm3inct_cll3_1920_160_75868179_2_9.zip for task hadcm3inct_cll3_1920_160_75868179_2 absent
28/08/2007 00:35:55|climateprediction.net|Output file hadcm3inct_cll3_1920_160_75868179_2_10.zip for task hadcm3inct_cll3_1920_160_75868179_2 absent
28/08/2007 00:35:55|climateprediction.net|Output file hadcm3inct_cll3_1920_160_75868179_2_11.zip for task hadcm3inct_cll3_1920_160_75868179_2 absent
28/08/2007 00:35:55|climateprediction.net|Output file hadcm3inct_cll3_1920_160_75868179_2_12.zip for task hadcm3inct_cll3_1920_160_75868179_2 absent
28/08/2007 00:35:55|climateprediction.net|Output file hadcm3inct_cll3_1920_160_75868179_2_13.zip for task hadcm3inct_cll3_1920_160_75868179_2 absent
28/08/2007 00:35:55|climateprediction.net|Output file hadcm3inct_cll3_1920_160_75868179_2_14.zip for task hadcm3inct_cll3_1920_160_75868179_2 absent
28/08/2007 00:35:55|climateprediction.net|Output file hadcm3inct_cll3_1920_160_75868179_2_15.zip for task hadcm3inct_cll3_1920_160_75868179_2 absent
28/08/2007 00:35:55|climateprediction.net|Output file hadcm3inct_cll3_1920_160_75868179_2_16.zip for task hadcm3inct_cll3_1920_160_75868179_2 absent
28/08/2007 00:36:57|climateprediction.net|Sending scheduler request: To fetch work
28/08/2007 00:36:57|climateprediction.net|Requesting 8640 seconds of new work, and reporting 1 completed tasks
28/08/2007 00:37:02|climateprediction.net|Scheduler RPC succeeded [server version 509]

---

Alan,

Unfortunately, your model has crashed (this one). The exit code \'22\' is a catch-all, but I haven\'t seen the specific final error before (i.e. \'Result too large\'). Perhaps someone else might know ...

A new model has been downloaded automatically, and you can continue with that. It is, however, possible to backup a CPDN model and restore it when a crash occurs: if the problem was PC-induced then the restored model will continue, if the problem was model-induced then it will crash again at the same point. Have a look here for instructions.

If you\'re making backups it\'s usually a good idea to prevent the automatic download of a new model - just press the \'no new tasks\' button in BOINC Manager.

Iain

Hi Alan

Just to add a couple of points to what Iain said (we must have been looking at your results at the same time). The models you\'ve had are listed here

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/results.php?hostid=530152

The recently crashed models all failed with a 22 code and the older ones with a -161 error code. Both codes can unfortunately result from several different problems. But they can\'t all result from flaws in the models you\'ve had. I\'m guessing that your computer isn\'t overclocked, so I think most of these crashes resulted from some event on your computer. Not necessarily always the same event.

The first line of defence is to start making backups as Iain said. I\'d also recommend a look at the project READMEs. Get to them through my signature. In the one called Running the model, check out the top tips.

Then in the README about avoiding model crashes, go to item #5 by Mike who outlines all the main precautions to help models reach the end.

---

Cpdn news

what type of \"event\" on my computer will cause these tasks to crash...

the copmuter is on 24/7 and to my knowledge as had no errors of crashes... viurs scans every day show nothing antispyware software show nothing...
so what do you think could cause these tasks to fail..

I take backups of CPDn every day

TIA
alan

---

It sounds like some sort of file corruption or buffer overflow problem, I\'ve never seen it before either.

...
BUFFIN: Read Failed: Result too large
BUFFIN: C I/O Error feof - Unit 68 - Return code = 16

Model crashed: umshell1.f:  STWORK  : I/O error - PP fixed length header
@
Sorry, too many model crashes! :-(

Curiously I found the same error at Seti via google:
http://setiweb.ssl.berkeley.edu/beta/result.php?resultid=1712761

This implies that it might be in the Berkeley code rather than the Oxford / Hadley code.

The feof() function looks for the end-of-file marker in a buffer.

---

I'm a volunteer and my views are my own.
News and Announcements and FAQ

The sorts of events that can cause a model crash are typically the things Mike mentions in that post in the Avoiding model crashes README. This doesn\'t mean your computer has a software or hardware defect. Crashes could be caused by

* Turning off the computer without exiting from boinc manager first and waiting for the system tray icon to disappear

* Doing graphics-intensive work on the computer eg playing games with the model running (usually leads to -107 error code which isn\'t what happened to your models)

* AV scanning without exiting from boinc first

* On some computers, running the screensaver all the time (one reason why most of us disable it)

* In a very few cases, Microsoft updates without exiting from boinc first

* Experimenting with the wrong boinc manager buttons (Reset, Abort, Detach)

* A power outage

* Not having enough RAM for the model you\'re running ie having less that 512Mb per running model

* Having a graphics card driver that needs an update from the manufacturer\'s website (typically leads to -107 or -1 error code)

* Computer overheating. You can download freebie temp monitors eg Everest if you suspect this.

* Overambitious overclocking causing the computer to become unstable in its calcualtions

* Faulty RAM card


If you want to check on stability and RAM which are hardware problems, the post on hardware maintenance and testing by UKNick in one of the READMEs gives excellent advice and links to testing programs that can be downloaded without charge.

I suspend my models and exit from boinc before I carry out my quite long list of regular computer \'housekeeping\' tasks. Even this isn\'t a total guarantee, so I still back up my models regularly. As far as I know, all the moderators have had to restore a backup of a crashed model at some time. The models do run very stably, but to guarantee that nothing untoward will happen during so many hours of computing is impossible.

---

Cpdn news

Hi Mike

Did you find that error in just one set of model crash messages or in several? If this could be a boinc code problem, maybe a boinc version update would help? Alan has 5.8.16. A newer version certainly wouldn\'t do any harm.

---

Cpdn news

I found it in one SETI workunit, and in a dozen or so CPDN workunits. The SETI message was identical to the CPDN message. 5.10.13 seems to be a reasonably stable version of the Boinc manager, so it sounds like a good idea.

---

I'm a volunteer and my views are my own.
News and Announcements and FAQ

I found it in one SETI workunit, and in a dozen or so CPDN workunits. The SETI message was identical to the CPDN message. 5.10.13 seems to be a reasonably stable version of the Boinc manager, so it sounds like a good idea.

Hi guys.. I have just tried to load the 5.10.13 version but I get a
Error 1327.Invaild Drive G:\\
on the BOINC Intsaller Information box

I don\'t have a drive \"G\" on my system...
any ideas
Thanks
Alan

---

Did you allow it to install in the default location, or did you try to select a different one yourself?

Did you allow it to install in the default location, or did you try to select a different one yourself?

Les just to the default...

---

If you look at My computer - C drive - Program files - is Boinc there in the list? And if you double-click on Boinc do you see a list of two or three folders and lots of files?

---

Cpdn news