climateprediction.net home page
Model won\'t start

Model won\'t start

Questions and Answers : Unix/Linux : Model won\'t start
Message board moderation

To post messages, you must log in.

AuthorMessage
old_user132927

Send message
Joined: 8 Dec 05
Posts: 4
Credit: 0
RAC: 0
Message 18557 - Posted: 21 Dec 2005, 15:12:48 UTC

I have tried to download and run climateprediction a few times on my Linux machine (redhat). It downloads everything fine, but never runs. It tells me \"Model hasn\'t started, please wait\".
Hours later, still nothing.
My work tab says its running, but no cpu total.
I have uninstalled everything and reinstalled three different times over the last two weeks, with the same result.
SETI works well, but have swithched to climateprediction without luck.
Any suggestions? I am using BOINC 5.2.13
It just won\'t start running!

David
ID: 18557 · Report as offensive     Reply Quote
old_user132927

Send message
Joined: 8 Dec 05
Posts: 4
Credit: 0
RAC: 0
Message 18558 - Posted: 21 Dec 2005, 15:43:45 UTC

As an additional debugging tool, I reran boinc. In my terminal window, I get the following:


Created shared memory region key = 72665 of size 569976 bytes
.so shmem return code = 0
Copying files for startup...
In pre_initialise_phase (part 1 of 3)
In initialise_phase (part 2 of 3)
In startup_phase (part 3 of 3)
Starting model ID sulphur_h067_000793375 Phase 1
Getting pthread attributes - retval=0
Setting pthread size (66560000 bytes) - retval=22
Waiting for model startup, this may take a minute...
Model timeout at 180.00 seconds
Preparing for restart...
Rewinding a model-day...
Starting model ID sulphur_h067_000793375 Phase 1
Getting pthread attributes - retval=0
Setting pthread size (66560000 bytes) - retval=22
Waiting for model startup, this may take a minute...



Hours later, nothing else happens. Does this help anybody debug my problem?
Thanks

ID: 18558 · Report as offensive     Reply Quote
Profile geophi
Volunteer moderator

Send message
Joined: 7 Aug 04
Posts: 2167
Credit: 64,403,322
RAC: 5,085
Message 18559 - Posted: 21 Dec 2005, 15:54:17 UTC

I just did a forum search for pthread. It looks like your post is the only one that\'s found. This must be a new error. Hopefully someone more fluent in Linux technicalities will come along and help figure this out.
ID: 18559 · Report as offensive     Reply Quote
Les Bayliss
Volunteer moderator

Send message
Joined: 5 Sep 04
Posts: 7629
Credit: 24,240,330
RAC: 0
Message 18561 - Posted: 21 Dec 2005, 17:01:23 UTC

I think there may be a clue in this line:
> Rewinding a model-day...

This usually means that the program has a problem, and is rewinding to a previous calcultion point to try again.
Except that there isn\'t one, because the model has only just started.

All of your many attempts have failed with an unknown error.
And I seem to remember Thyme Lawn posting that the error \'161\' shown on them just means that when BOINC tried to upload the failure data to the server, there weren\'t any files to upload.
Which brings us back to the usual questions about stability and software conflicts.
And I don\'t know enough about Linux to know where to begin.

I\'ll point you to a couple of posts about these things, and hope.
<a href=\"http://www.climateprediction.net/board/viewtopic.php?t=2124\"> Here</a> and <a href=\"http://www.climateprediction.net/board/viewtopic.php?t=2126\"> here.</a>

ID: 18561 · Report as offensive     Reply Quote
Profile astroWX
Volunteer moderator

Send message
Joined: 5 Aug 04
Posts: 1496
Credit: 95,522,203
RAC: 0
Message 18580 - Posted: 21 Dec 2005, 20:42:51 UTC
Last modified: 21 Dec 2005, 22:02:04 UTC

Actually, its an old problem, one I\'ve also had.
http://climateapps2.oucs.ox.ac.uk/cpdnboinc/forum_thread.php?id=3187

I\'m searching for more info -- and will look at the Sulphur Beta BB, as well.

Edit. Took awhile and I didn\'t find much. This was a problem in Beta v. 4.17 and it didn\'t affect all Linux versions equally.

I hope it doesn\'t require a new version to fix!

Posted by Tolu on the SC Beta BB after we reported the problem:
This was a problem in the boinc api with timer handling .
David.A fixed this this morning and i\'ve released a new app version for linux. 4.18. This would also resolve the initial linux failures.


(Tolu\'s post dealt with stoppage of Terminal Window logging and the failure-to-start issue.)
"We have met the enemy and he is us." -- Pogo
Greetings from coastal Washington state, the scenic US Pacific Northwest.
ID: 18580 · Report as offensive     Reply Quote
Les Bayliss
Volunteer moderator

Send message
Joined: 5 Sep 04
Posts: 7629
Credit: 24,240,330
RAC: 0
Message 18583 - Posted: 21 Dec 2005, 22:00:02 UTC

OK. Leave it with you.

ID: 18583 · Report as offensive     Reply Quote
old_user132927

Send message
Joined: 8 Dec 05
Posts: 4
Credit: 0
RAC: 0
Message 18615 - Posted: 22 Dec 2005, 13:48:57 UTC

I am using sulphur version 4.22 without success. Still won\'t start.
ID: 18615 · Report as offensive     Reply Quote
Profile astroWX
Volunteer moderator

Send message
Joined: 5 Aug 04
Posts: 1496
Credit: 95,522,203
RAC: 0
Message 18631 - Posted: 22 Dec 2005, 20:38:24 UTC - in response to Message 18615.  
Last modified: 22 Dec 2005, 20:48:41 UTC

I am using sulphur version 4.22 without success. Still won\'t start.


Understood; no one is using the old versions (though I have some Beta Runs suspended in favor of Spinup). The point of the post was that the \"fix\" was a new client.

Geophi had the failure with Mandrake and one other implementation. Mine were with SuSE 9.0 and 9.1. As geophi pointed out in a PM, \"pthread\" wasn\'t reported then, so this may not be the same failure. I focused on the \"won\'t start\" aspect of the problem.

As was true in Beta and the thread noted below, my machines got cold awaiting a new release. (I don\'t run other boinc projects.) Bottom line, I can only offer empathy, not help.

Edit: If you haven\'t tried reboot and Reset project (in Projects Tab) to start from the beginning, you have nothing to lose.

By the way, if all four of the machines listed in Your account are physically the same box, they can be merged.
"We have met the enemy and he is us." -- Pogo
Greetings from coastal Washington state, the scenic US Pacific Northwest.
ID: 18631 · Report as offensive     Reply Quote
old_user132927

Send message
Joined: 8 Dec 05
Posts: 4
Credit: 0
RAC: 0
Message 18659 - Posted: 23 Dec 2005, 14:44:27 UTC - in response to Message 18631.  
Last modified: 23 Dec 2005, 14:45:33 UTC

Hello AstroWX,
Yes, I have tried (multiple times) to reset the project and start over. Each time, the same thing. Oh well....
Thanks for your empathy.
ID: 18659 · Report as offensive     Reply Quote
Profile Ananas
Volunteer moderator

Send message
Joined: 31 Oct 04
Posts: 336
Credit: 3,316,482
RAC: 0
Message 18690 - Posted: 24 Dec 2005, 5:01:24 UTC
Last modified: 24 Dec 2005, 5:03:25 UTC

If you edit client_state.sah and reduce the values on_frac, connected_frac and active_frac in the time_stats section to maybe 0.1, it will probably stop assigning sulphur models and give you one of the \"normal\" ones instead. Maybe your PC is more happy with that.

This is not really a solution of course - but maybe an intermediate workaround until a version is out that works better on your box
ID: 18690 · Report as offensive     Reply Quote
Les Bayliss
Volunteer moderator

Send message
Joined: 5 Sep 04
Posts: 7629
Credit: 24,240,330
RAC: 0
Message 18693 - Posted: 24 Dec 2005, 5:37:25 UTC

There aren\'t any slab models now, just sulphur.
Some computers just won\'t work with BOINC no matter what you do. :(

ID: 18693 · Report as offensive     Reply Quote
Helmer Bryd

Send message
Joined: 16 Aug 04
Posts: 147
Credit: 7,748,561
RAC: 8,366
Message 18718 - Posted: 25 Dec 2005, 1:25:21 UTC

Guess this thread is somewhat related:
http://www.climateprediction.net/board/viewtopic.php?t=3402
ID: 18718 · Report as offensive     Reply Quote
old_user21637

Send message
Joined: 28 Sep 04
Posts: 36
Credit: 268,150
RAC: 0
Message 19053 - Posted: 5 Jan 2006, 17:16:56 UTC - in response to Message 18718.  


The problem seems to have been identified. It seems to be an incompatibility with GLIBC windows library versions.

Please see thread:

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/forum_thread.php?id=3822

I have put together a patch that might help fix the problem, until a new version is released.

Cheers,
Stefan.


ID: 19053 · Report as offensive     Reply Quote

Questions and Answers : Unix/Linux : Model won\'t start

©2024 climateprediction.net