climateprediction.net home page
Completion

Completion

Message boards : climateprediction.net Science : Completion
Message board moderation

To post messages, you must log in.

1 · 2 · Next

AuthorMessage
MW

Send message
Joined: 11 Dec 06
Posts: 46
Credit: 5,034,990
RAC: 0
Message 30677 - Posted: 24 Sep 2007, 9:42:16 UTC

Just completed a model - status says \"Ready to report\". When can I switch off?
ID: 30677 · Report as offensive     Reply Quote
Profile Pooh Bear 27
Avatar

Send message
Joined: 5 Feb 05
Posts: 465
Credit: 1,914,189
RAC: 0
Message 30678 - Posted: 24 Sep 2007, 10:06:04 UTC

BOINC will report it automatically. In CPDN case it will usually do it within the next 24 hours.

ID: 30678 · Report as offensive     Reply Quote
Les Bayliss
Volunteer moderator

Send message
Joined: 5 Sep 04
Posts: 7629
Credit: 24,240,330
RAC: 0
Message 30680 - Posted: 24 Sep 2007, 16:53:21 UTC


Or you can Exit BOINC, shutdown the computer, and wait at least 6 hours.
Then start the computer, start BOINC and see if it \'Reports\'.
If not, click the Update button in the Projects tab.

DO NOT do this too early!
The delay is to allow the data to be moved from the upload server to the data server. If the Report message gets there too soon, it will be rejected because the data server won\'t know about the model\'s final data. And then the model will never show as being completed.

Patience is needed for this.

ID: 30680 · Report as offensive     Reply Quote
MW

Send message
Joined: 11 Dec 06
Posts: 46
Credit: 5,034,990
RAC: 0
Message 30708 - Posted: 26 Sep 2007, 7:34:55 UTC - in response to Message 30680.  

Thank you, Pooh Bear and Les. It reported within the six hours.
I have now resurrected a back-up of another model from March this year; the trickles are being accepted. Does this indicate that this model has not been reassigned to another computer? If so, I will keep going with it; if not, I\'ll abort.
ID: 30708 · 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 30709 - Posted: 26 Sep 2007, 7:47:01 UTC


I\'m not sure which model it is, is it this one (crashed on the 24th March):

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/workunit.php?wuid=5910633

If so, it looks like another model was issued, but that new model was never heard of again (lost, in other words).



I'm a volunteer and my views are my own.
News and Announcements and FAQ
ID: 30709 · Report as offensive     Reply Quote
MW

Send message
Joined: 11 Dec 06
Posts: 46
Credit: 5,034,990
RAC: 0
Message 30711 - Posted: 26 Sep 2007, 8:14:47 UTC
Last modified: 26 Sep 2007, 8:27:15 UTC

No, Mike. 5910633 crashed around (model year) 2000 due to negative pressure. And it was a b****y looper. The one now crunching is 5979988 which crashed on 1st. April (model year 1926), although it\'s one of the hottest, so I don\'t have high expectations of its stability. Correction - I originally aborted it because of my expectation (and frustration at the time, having not completed a model in over twelve months and apparently losing over 60,000 credits on the BBC-side of things!)
ID: 30711 · Report as offensive     Reply Quote
Profile Iain Inglis

Send message
Joined: 9 Jan 07
Posts: 467
Credit: 14,549,176
RAC: 317
Message 30715 - Posted: 26 Sep 2007, 10:22:54 UTC

When your model from the 5979988 work unit finishes, you\'ll get some alarming messages about the results not being accepted: don\'t be put off by those. The model is already trickling and if those trickles continue to be recorded on the Web site (and the associated Zip files are uploaded) then that\'s what matters. Credits will accrue as usual even though the Web site thinks the model has crashed.
ID: 30715 · Report as offensive     Reply Quote
Profile mo.v
Volunteer moderator
Avatar

Send message
Joined: 29 Sep 04
Posts: 2363
Credit: 14,611,758
RAC: 0
Message 30718 - Posted: 26 Sep 2007, 15:40:26 UTC
Last modified: 26 Sep 2007, 15:58:06 UTC

MW, this workunit crashed for its other owner who has had a lot of trouble with 22 errors but is managing with his current (different) model so far.

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/workunit.php?wuid=5979988

So you\'re the only person at the moment with a chance of completing it..... If I were you, I\'d keep the screensaver disabled to run the model as fast as possible and make a boinc folder backup at least weekly to maximise the chance of completing it. As we learn from bitter experience, with such long models the probability that something will go wrong before the end is quite high.

So what happened to your BBC credits? All gone, I see, which shouldn\'t happen. Did you use the same email address for BBC and cpdn or a different one?
Cpdn news
ID: 30718 · Report as offensive     Reply Quote
Profile mo.v
Volunteer moderator
Avatar

Send message
Joined: 29 Sep 04
Posts: 2363
Credit: 14,611,758
RAC: 0
Message 30719 - Posted: 26 Sep 2007, 16:15:53 UTC
Last modified: 26 Sep 2007, 16:18:54 UTC

There are quite a few crunchers called MW, but I think this is your cpdn record on BoincStats:

http://boincstats.com/search/all_projects.php?cpid=92a9721f4396305afe40b5172862d3f8

and this is your separate BBC stats record:

http://boincstats.com/search/all_projects.php?cpid=7b8eb1d5892c11ea8703c750faedf546

Each account has a different CPID. If you change the email address of one project account so it\'s the same as the other, I think the two stats records should merge into one.
Cpdn news
ID: 30719 · Report as offensive     Reply Quote
MW

Send message
Joined: 11 Dec 06
Posts: 46
Credit: 5,034,990
RAC: 0
Message 30734 - Posted: 27 Sep 2007, 9:43:05 UTC

Yep, that certainly seems to be me! AFAIK, my e-mail address is the same on both sites; I recall I jumped the BBC ship in total frustration (for the first time) having got to 2069 before a computer disaster wiped everything. I then joined CPDN and sought advice at the time, Nov/Dec 2006, as to how I could get my BBC credits, and antipodean Les tried to assist, but my computer illiteracy prevented further progress in the matter.
With regard to hadcm3pbb_bp04_05819695, it would seem that the second \"owner\" lost it due to negative pressure in its first year (no trickles), whereas my attempt at it got it, initially, to 1926, although it crashed yesterday with \"computational error\" in 1928. I don\'t know how to find out whether that was a machine or a model fault. I\'m running it again from 1926, but if it dies again in 1928, that\'ll be it.
By the way, you may recall me as having drawn attention last year to at least five crunchers crunching the pre-Easter 2006 models.
ID: 30734 · Report as offensive     Reply Quote
Profile mo.v
Volunteer moderator
Avatar

Send message
Joined: 29 Sep 04
Posts: 2363
Credit: 14,611,758
RAC: 0
Message 30736 - Posted: 27 Sep 2007, 10:13:48 UTC

I assume you\'ve only ever had the one computer crunching for BBC and cpdn?

I think you should still go into your accounts for BBC and cpdn to check whether the email addresses really are the same. If they\'re different, change one.

After that (and whether or not you need to change an addy), if your computer is still attached to BBC, highlight BBC in the Projects tab and click Update.


Cpdn news
ID: 30736 · Report as offensive     Reply Quote
MW

Send message
Joined: 11 Dec 06
Posts: 46
Credit: 5,034,990
RAC: 0
Message 30737 - Posted: 27 Sep 2007, 10:57:17 UTC

Yes, just the one computer. I added memory, but that was during the BBC run, and is acknowledged on the BBC site. The e-mail addresses are the same. However, the Domain Name and Name of the computer are different, BBC to CPDN; does this matter? I don\'t know how to change one or the other.
I\'m no longer attached to the BBC site, and don\'t know how to attach, even if this is still possible. That was the problem when Les tried to help.
ID: 30737 · 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 30738 - Posted: 27 Sep 2007, 12:15:40 UTC


I think you can still attach to it using the attach wizard (http://bbc.cpdn.org). When it asks you for the username etc, make sure you select \'existing user\'.

The email address must be identical on both projects. Once you\'re attached to both, click \'update\' a few times on each project.
I'm a volunteer and my views are my own.
News and Announcements and FAQ
ID: 30738 · Report as offensive     Reply Quote
MW

Send message
Joined: 11 Dec 06
Posts: 46
Credit: 5,034,990
RAC: 0
Message 30740 - Posted: 27 Sep 2007, 14:10:13 UTC

Mike, the link you give doesn\'t seem to help. It merely shows that the BBC expt. is closed to newcomers and directs them to CPDN, unless I\'m being thicker than normal. Any ideas?
ID: 30740 · Report as offensive     Reply Quote
Profile mo.v
Volunteer moderator
Avatar

Send message
Joined: 29 Sep 04
Posts: 2363
Credit: 14,611,758
RAC: 0
Message 30741 - Posted: 27 Sep 2007, 14:50:47 UTC
Last modified: 27 Sep 2007, 14:56:22 UTC

In your boinc manager, click Tools, then select Attach to project. Up jumps the project attachment wizard. You\'ll see a list of projects to attach to and BBC won\'t be there. So instead, type or paste the address Mike gave into the box. You\'ll then be told that BBC isn\'t accepting new members and you can only attach if you have an existing account. So say yes, you\'re an existing member. You\'ll have to type in your BBC email address and password to prove it.

BBC should then appear in your boinc manager Projects tab. As it will immediately try to fetch work that\'s no longer available, you\'ll have to select it (again in the Projects window) and click No new work.

Up to this point I know everything works - I\'ve just done it.

Then allow network activity for boinc, select BBC and click Update. More than once to make sure the BBC server gets your message. I would also click Update for cpdn. I hope you\'ll then see a boinc manager message to say you\'re being given a new BBC CPID. Or a new cpdn one, it doesn\'t really matter which.

It may take a couple of days for the New Unified You to show up on the stats sites.

Cpdn news
ID: 30741 · Report as offensive     Reply Quote
MW

Send message
Joined: 11 Dec 06
Posts: 46
Credit: 5,034,990
RAC: 0
Message 30744 - Posted: 27 Sep 2007, 16:34:33 UTC

Thanks, mo.v, I\'ll give it a whirl.
ID: 30744 · Report as offensive     Reply Quote
MW

Send message
Joined: 11 Dec 06
Posts: 46
Credit: 5,034,990
RAC: 0
Message 32193 - Posted: 16 Jan 2008, 11:10:44 UTC - in response to Message 30718.  

MW, this workunit crashed for its other owner who has had a lot of trouble with 22 errors but is managing with his current (different) model so far.

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/workunit.php?wuid=5979988

So you\'re the only person at the moment with a chance of completing it..... If I were you, I\'d keep the screensaver disabled to run the model as fast as possible and make a boinc folder backup at least weekly to maximise the chance of completing it. As we learn from bitter experience, with such long models the probability that something will go wrong before the end is quite high.

So what happened to your BBC credits? All gone, I see, which shouldn\'t happen. Did you use the same email address for BBC and cpdn or a different one?

I have nursed this along to mid-2036, but has now crashed twice. I\'m running another back-up, but, boy, does it look hot! How can I tell if it\'s now non-viable?
ID: 32193 · 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 32195 - Posted: 16 Jan 2008, 12:46:31 UTC


The main thing is whether it is crashing at the same point, or a different point. if it crashes more than once at the same point, it is non-viable.

In this case we can\'t tell viability from the error log on the website because it just reflects the first reported error (aborted).

I'm a volunteer and my views are my own.
News and Announcements and FAQ
ID: 32195 · Report as offensive     Reply Quote
Profile mo.v
Volunteer moderator
Avatar

Send message
Joined: 29 Sep 04
Posts: 2363
Credit: 14,611,758
RAC: 0
Message 32199 - Posted: 16 Jan 2008, 16:28:29 UTC

The graph is so hot that I\'m posting it for all to see. This could be the hottest model ever!


If it does crash again, keep a note of the model crash date so you can compare this with any future crash dates.

Unusual outlying results like yours are also useful for statistical purposes, so thanks for making the effort to keep this model going.
Cpdn news
ID: 32199 · 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 32200 - Posted: 16 Jan 2008, 17:00:23 UTC



Quite an interesting set of parameters on that model...

* High humidity (so warmer)

* High entrainment coefficient (so sensitive to high CO2, hence warmer in the forecast phase)

* Reduced sulphate coefficient (global cooling due to particulates is reduced, hence warmer)


I'm a volunteer and my views are my own.
News and Announcements and FAQ
ID: 32200 · Report as offensive     Reply Quote
1 · 2 · Next

Message boards : climateprediction.net Science : Completion

©2024 climateprediction.net