climateprediction.net home page
Premature termination...

Premature termination...

Message boards : climateprediction.net Science : Premature termination...
Message board moderation

To post messages, you must log in.

AuthorMessage
old_user109453

Send message
Joined: 21 Nov 05
Posts: 3
Credit: 60,525
RAC: 0
Message 24221 - Posted: 6 Sep 2006, 15:16:37 UTC

Hi,

I have been plugging along with a WU for several months (the hadcm3b one) and it had finally reached almost 25% complete.

All of a sudden, I get a message saying that the computation was finished, and it downloaded a new unit and started work on it.

Is this normal? I was expecting months more of slogging along - the finish due date was sometime in January 2008! How can it go from 25% to finished overnight?

Thanks in advance,

Phil.
ID: 24221 · Report as offensive     Reply Quote
Profile MikeMarsUK
Volunteer moderator
Avatar

Send message
Joined: 13 Jan 06
Posts: 1498
Credit: 15,613,038
RAC: 0
Message 24223 - Posted: 6 Sep 2006, 19:15:51 UTC

Hi,

It looks like your model crashed with a \'-161\' error. This doesn\'t actually tell us anything useful, it just means \'the model finished, and there is nothing to upload\'. The topmost post on the \'Windows Q&A\' forum discusses this particular error in more depth.

The fortunate thing is that the crash happened at 1961 - at the years 1960, 2000 and 2040 a \'restart dump\' is uploaded to the servers. This means that in the future, when the software is written, a new model can be created from yours which runs from 1960-2080.

What I would advise is doing a backup at intervals (I do them weekly), this can be as simple as the following:

* Right-click on the boinc icon, \'exit\'
* Navigate to c:\\program files\\ using Windows explorer or My Computer
* Right-click on \\boinc\\, \'copy\', \'paste\'
* (If you got a \'files in use\' error, then reboot, and repeat the above steps)
* Then run the manager again to continue processing.
I'm a volunteer and my views are my own.
News and Announcements and FAQ
ID: 24223 · Report as offensive     Reply Quote
old_user109453

Send message
Joined: 21 Nov 05
Posts: 3
Credit: 60,525
RAC: 0
Message 24232 - Posted: 7 Sep 2006, 16:27:38 UTC - in response to Message 24223.  

Hi,

It didn\'t give me any kind of error message. According to the messages listed, it just completed. I have tried to cut&paste the messages around the completion, but it won\'t let me... It had finished uploading a trickle-up message, then it says resuming result hadcm... The next line says that \"computation for result hadcm... finished\". Then it sent a scheduler request for more work.

Also, are you suggesting I \'copy\' and \'paste\' the actual boinc directory? I have my hard-drive partitioned, so my program is on another disk, but that is what it looks like you are suggesting. What do you paste it into?

Thanks,

Phil.




Hi,

It looks like your model crashed with a \'-161\' error. This doesn\'t actually tell us anything useful, it just means \'the model finished, and there is nothing to upload\'. The topmost post on the \'Windows Q&A\' forum discusses this particular error in more depth.

The fortunate thing is that the crash happened at 1961 - at the years 1960, 2000 and 2040 a \'restart dump\' is uploaded to the servers. This means that in the future, when the software is written, a new model can be created from yours which runs from 1960-2080.

What I would advise is doing a backup at intervals (I do them weekly), this can be as simple as the following:

* Right-click on the boinc icon, \'exit\'
* Navigate to c:\\program files\\ using Windows explorer or My Computer
* Right-click on \\boinc\\, \'copy\', \'paste\'
* (If you got a \'files in use\' error, then reboot, and repeat the above steps)
* Then run the manager again to continue processing.


ID: 24232 · Report as offensive     Reply Quote
Profile MikeMarsUK
Volunteer moderator
Avatar

Send message
Joined: 13 Jan 06
Posts: 1498
Credit: 15,613,038
RAC: 0
Message 24233 - Posted: 7 Sep 2006, 19:51:41 UTC

Into the \\program files\\ directory - it\'s simply making a copy of the whole directory (will appear as \\copy of boinc\\).

There are any numbers of ways of doing backups, but this is what it boils down to.
I'm a volunteer and my views are my own.
News and Announcements and FAQ
ID: 24233 · Report as offensive     Reply Quote
old_user109453

Send message
Joined: 21 Nov 05
Posts: 3
Credit: 60,525
RAC: 0
Message 24235 - Posted: 7 Sep 2006, 21:30:18 UTC - in response to Message 24233.  

Ok, thanks. Just never heard of using the clipboard to back something up like that is all...

Still missing Unix... *sigh*

Phil.


Into the \\program files\\ directory - it\'s simply making a copy of the whole directory (will appear as \\copy of boinc\\).

There are any numbers of ways of doing backups, but this is what it boils down to.


ID: 24235 · Report as offensive     Reply Quote
old_user39169

Send message
Joined: 27 Jan 05
Posts: 3
Credit: 730,315
RAC: 0
Message 24291 - Posted: 15 Sep 2006, 0:29:11 UTC

same thing has happened to me today too.

these are the corresponding messages:

12/09/2006 22:30:17|climateprediction.net|hadcm3lb not responding to screensaver, exiting
12/09/2006 22:30:20|climateprediction.net|Unrecoverable error for result hadcm3lbm_3gm8_05117645_1 ( - exit code -1 (0xffffffff))

boinc has now downloaded a new unit and is back to just over 1%. i had left it earlier today with about 15% work done...
the even more annoying thing is that this now has happened the 3rd time already. i\'m about to give up on this project...

thanks,
o
ID: 24291 · Report as offensive     Reply Quote
Les Bayliss
Volunteer moderator

Send message
Joined: 5 Sep 04
Posts: 7629
Credit: 24,240,330
RAC: 0
Message 24292 - Posted: 15 Sep 2006, 2:15:01 UTC

I think that exit code 1 is when the program doesn\'t get time to shut down the model before Windows exits.

You need to manually exit from the program first, and THEN shut down Windows.
Better still, in the menu, Suspend BOINC, wait until the model status in Tasks SAYS Suspended, and then exit.

The huge number of failures on your laptop may be due to the progarm, (or perhaps it\'s BOINC, I forget, now) not liking to \"hibernate\".

ID: 24292 · Report as offensive     Reply Quote

Message boards : climateprediction.net Science : Premature termination...

©2024 climateprediction.net