I noticed that I am having problems with this WU. I am on 2nd phase half way thru that one too. I too noticed that my time on that WU was increasing more than it ever has. I have complete many complete projects.
I am reluctant to abort this too. Blue world too.

---

SIIFRED

If you don\'t abort it, it may take another year to complete it.
And the data will be of questionable value.

---

Backups: Here

I know we learn to love our long models and are reluctant to abort them, but these \'iceworlds\' contain a defect and we cannot repair them.

---

Cpdn news

I know we learn to love our long models and are reluctant to abort them, but these \'iceworlds\' contain a defect and we cannot repair them.

If you don\'t abort it, it may take another year to complete it.
And the data will be of questionable value.

please ... kindly forgive me ... I\'m still learning about climateprediction.net

BOINC 6.10.18
Windows XP: Professional x86 Edition, Service Pack 3
my computer

I\'m crunching the following ... 24/7

UK Met Office HADSM3 Mid-Holocene
hadsm3mh_kvhx_006489753

UK Met Office HADSM3 Mid-Holocene
hadsm3mh_krg0_006484500

UK Met Office HADSM3 Mid-Holocene
hadsm3mh_kr54_006484108

UK Met Office HadSM3 Slab Model
hadsm3fub_k2w4_006460926

How do I tell if I have one of these ? - -> \" these \'iceworlds\' contain a defect and we cannot repair them\" ?

Should I abort any of these ?

I\'m still learning about climateprediction.net

thanks in advance.

How do I tell if I have one of these ?

Your computer is running Windows on an Intel processor, which means that an iceworld would slow down dramatically - the interval between trickles would stretch from a few hours to a week or more. The graphics display for temperature would also be blue, indicating uniform cold temperatures, which is why the model is then called an \'iceworld\'.

Though the models you list are running quite slowly, they have always run slowly - there is no slowdown and they are therefore not iceworlds.

\" these \'iceworlds\' contain a defect and we cannot repair them\" ?

Any iceworld that starts in the first phase will continue as an iceworld until the end of the model, which is three phases for a slab (HADSM3) model and four phases for a mid-holocene (HADSM3MH) model. An iceworld that starts in any phase other than the first will revert to normal processing at the start of the next phase (though it could freeze again).

Since iceworlds on Windows/Intel run so slowly, the advice is to abort them so that the computer can get on with processing another model that runs at normal speed.

Since iceworlds on Windows/Intel run so slowly, the advice is to abort them so that the computer can get on with processing another model that runs at normal speed.

aha... OK ... Thank you Iain.
Byron.

please ... kindly forgive me again ... I\'m still learning about climateprediction.net ...
one more question ...
which are ... the another models ... that runs at normal speed ... so that I can set my preferences to down load ... them ?

<Edit> ... or which models do you recomend ... or advise me to ... set my preferences to ...

also ... I just now notice the links in Les Bayliss\'s ... Signiture preferences to ?

Hints & tips: Here
Backups:Here

I\'ll read up on those links to educate my self about climateprediction.net
also I have subcribed to this thread.

thanks in advace.
Byron

Since iceworlds on Windows/Intel run so slowly, the advice is to abort them so that the computer can get on with processing another model that runs at normal speed.

aha... OK ... Thank you Iain.
Byron.

please ... kindly forgive me again ... I\'m still learning about climateprediction.net ...
one more question ...
which are ... the another models ... that runs at normal speed ... so that I can set my preferences to down load ... them?

also ... I just now notice the links in Les Bayliss\'s ... Signiture

Hints & tips: Here
Backups:Here

I\'ll read up on those links to educate my self about climateprediction.net
also I have subcribed to this thread.

thanks in advace.
Byron

Right now only the HADSM3 work units are available. The Hadam3p should be back “soon.”

---

One additional hadsm3 ice world model to report:

WorkUnit 6678256, Task ID 10246960.

Machine is Intel Xenon CPU, Windows XP SP3. Once again I find that the Temperature graphic is blue, with no clouds except around the Antarctic coast. Also the model\'s time to completion is rising instead of falling, and for quite some time now the model has not been uploading trickles.

-- Jim

Well done for recognising the problem, Jim.

I can only conclude from looking at all the tasks in your workunit that the majority of members never look at the graphics, or don\'t realise that they should ask about abnormal graphics, or know there\'s a problem but don\'t know what to do, or set & forget without ever checking anything.

I\'m unhappy about this situation because a lot of computer time and electricity are being wasted by those models. I\'ll be reporting it to the programmers.

---

Cpdn news

I can only conclude from looking at all the tasks in your workunit that the majority of members never look at the graphics, or don\'t realise that they should ask about abnormal graphics, or know there\'s a problem but don\'t know what to do, or set & forget without ever checking anything.

I confess that of late I myself have been in the \"never look at the graphics / set it & forget it\" work mode too. But I have five similar low-end dual Xenon CPU servers (cheap eBay acquisitions) I\'m using as high throughput workstations for data analysis (and for CPDN when not engaged in that activity), and yesterday I noticed that the \"average work done\" tally on one of the machines was roughly half that of the others, suggesting one of the pair of CPDN models being run on that machine hadn\'t been uploading trickles for some time. But anyway, looking at one\'s CPDN statistics might be another way for members to notice a problem, since ice world models generally ceases earning credit when they stop uploading trickles, yes?

Are you still in the habit of sending private e-mails to people still running models that have become iceworlds, suggesting that they abort and/or restore it from a pre-ice-world backup to capture ice-world-entry data for Iain ?

Regards,

-- Jim

Other mods or I used to send PMs to the other crunchers in the WU with the same OS, whether they were already or potentially affected. However, I\'ve never received a single reply to any of these PMs; the experience of other mods must be the same because one said on this forum that it\'s a waste of time so he no longer sends PMs.

This is because email notification of PMs is turned off by default and hardly any members enable it. That\'s the default Boinc setting. It\'s supposed to protect members\' privacy. You only get emails if you\'ve actively asked for them.

The mods can\'t send emails to members, only PMs. We can\'t see the database or accounts so we can\'t see members\' email addresses.

I\'ve suggested to the programmers that if they don\'t want to go against the default Boinc policy for PMs CPDN could perhaps have a special iceworld email. The mods would list the IDs of the affected computers and Milo would send emails to their owners.

---

Cpdn news

Hi. Think, but not sure, I have this iceworld problem.
My model is at
http://climateapps2.oucs.ox.ac.uk/cpdnboinc/show_user.php?userid=606429
It\'s in Phase 2 of 3, the globe is totally blue and I seem to be making no progress. For the past 2 months I\'ve been sitting on 34.925% completion, even though CPU time has now reached 850 hours (and climbing).
Current timestep is 12,386, of 259,248
10.90 s/TS, whatever that is.
hadsm3 version 6.07
Model time 01:00, Date 19/08/1826
Processor Intel Celeron CPU 2.40GHz, 1.5GB of RAM, absolutely standard (ie. no overclocking), WIndows XP Pro Service Pack 3

I suspect I need to abort, but even dumber I can\'t see how to.
Thanks,
Michael.

Hi. Think, but not sure, I have this iceworld problem.
My model is at
http://climateapps2.oucs.ox.ac.uk/cpdnboinc/show_user.php?userid=606429
It\'s in Phase 2 of 3, the globe is totally blue and I seem to be making no progress. For the past 2 months I\'ve been sitting on 34.925% completion, even though CPU time has now reached 850 hours (and climbing).
Current timestep is 12,386, of 259,248
10.90 s/TS, whatever that is.
hadsm3 version 6.07
Model time 01:00, Date 19/08/1826
Processor Intel Celeron CPU 2.40GHz, 1.5GB of RAM, absolutely standard (ie. no overclocking), WIndows XP Pro Service Pack 3

I suspect I need to abort, but even dumber I can\'t see how to.
Thanks,
Michael.

PS. Also can\'t see where to turn on PM\'s

Oh, and coincidentally, it has just done this!
hadsm3_graphics_6.07_windows_intelx86.exe has encountered a problem and needs to close. We are sorry for the inconvenience.

I know I\'ve seen this at least once before, but it could be happening regularly for all I know - screen not normally visible.

Hi Michael, welcome to the forum.

PMs In Community preferences on your Account page, select 1 of 3 options for email.

10.90 s/TS, whatever that is. seconds per Time Step. For your computer, and that model, it\'s taking 10.9 seconds to process 1 Time Step. (Certain number of boexs across the planet X a certain number of boxes down the planet X a certain number of boxes vertically into the air. Plus, for some models, some boxes down into the ocean. See the science section at the front of this forum.)

Aborting Click on the name of the model in the Tasks tab, and then click on the Abort button to the left.

Encountered a problem Sometimes happens. Could be that something has interrupted the model at a critical point.

---

Backups: Here

Hi Michael, welcome to the forum.

.....

Thanks heaps, for both the welcome and the patient explanation/s.
Bit the bullet and aborted. I see it\'s another hadsm3 task - hopefully will be more productive.
Cheers,
Michael.

You have a recent version of Boinc so I assume you can see the graphics using the View graphics button. Have a quick look at your globe every couple of days if you can to check it\'s still normal with all the colours. If you suddenly see just one colour, which will be the default, this is the quickest diagnosis.

---

Cpdn news

I have aborted this task:
http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=10547555
I\'m not good at all the technical details, but I did see that my planet was completely blue.

The model was running on an iMac Core 2 Duo 3.06 EE, not overclocked, running the previous BOINC version (.18) Vista.
The second model on the other CPU core is still running fine.

I don\'t know exactly when it went wrong, but I\'d thought it would be a good idea anyway to let you know this.

I have aborted this task:
http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=10547555
I\'m not good at all the technical details, but I did see that my planet was completely blue.

The model was running on an iMac Core 2 Duo 3.06 EE, not overclocked, running the previous BOINC version (.18) Vista.
The second model on the other CPU core is still running fine.

I don\'t know exactly when it went wrong, but I\'d thought it would be a good idea anyway to let you know this.

That model seems to have been running on Intel/Vista rather than Mac, so you have done the right thing to abort it. It would have slowed down greatly. It would have recovered at the start of phase 4 and probably completed - but it might have taken you a month to get to end of phase 3!

Next week Milo will implement a CPDN email that can be sent to members who have iceworlds but don\'t know about the problem. The email will contain a link to the forum so they can ask for advice. This thread is very long so I have started a new iceworld discussion thread here. Pease report iceworlds there.

---

Cpdn news