climateprediction.net home page
Iceworlds & Slowdowns hadsm3/mh - Closed - Discussion

Iceworlds & Slowdowns hadsm3/mh - Closed - Discussion

Message boards : Number crunching : Iceworlds & Slowdowns hadsm3/mh - Closed - Discussion
Message board moderation

To post messages, you must log in.

Previous · 1 . . . 10 · 11 · 12 · 13 · 14 · 15 · Next

AuthorMessage
old_user13951

Send message
Joined: 6 Sep 04
Posts: 12
Credit: 1,767,663
RAC: 0
Message 36488 - Posted: 26 Mar 2009, 17:20:55 UTC


I think I have eight of these bad boys:

7714560
7714126
7703437
7703654
7703204
7703491
7714497
7703217

Should I abort them?
ID: 36488 · Report as offensive
wateroakley

Send message
Joined: 6 Aug 04
Posts: 185
Credit: 27,083,655
RAC: 6,161
Message 36489 - Posted: 26 Mar 2009, 18:04:52 UTC

They all appear to be very slow processing, some for up to 3 months. Presumably they are also blue ice-worlds? Your computers have succesfully run other models. I\'d suggest you abort the slow running models. Good luck with the new ones.
ID: 36489 · Report as offensive
old_user13951

Send message
Joined: 6 Sep 04
Posts: 12
Credit: 1,767,663
RAC: 0
Message 36490 - Posted: 26 Mar 2009, 18:51:29 UTC - in response to Message 36489.  

They all appear to be very slow processing, some for up to 3 months. Presumably they are also blue ice-worlds? Your computers have succesfully run other models. I\'d suggest you abort the slow running models. Good luck with the new ones.


They do all display totally Blue on graphics.
ID: 36490 · Report as offensive
RichardRodd

Send message
Joined: 16 Mar 06
Posts: 28
Credit: 3,219,100
RAC: 0
Message 36507 - Posted: 27 Mar 2009, 12:45:38 UTC

Greetings.

I\'ve just spotted a couple of very slow-running models. I should probably abort - can someone please advise...

1. http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=7673807
Current timestep = 49417 of 259248
s/TS = 56.6

2. http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=7731951
Current timestep = 39775 of 259248
s/TS = 3.64

Neither are blue / ice worlds, but the to-completion time is rising, so maybe they\'re turning blue...
It\'s an Intel machine runnig WinXP, not overclocked and has, prior to these two models, turned in a few complete runs.

Thanks - Richard
ID: 36507 · Report as offensive
Profile geophi
Volunteer moderator

Send message
Joined: 7 Aug 04
Posts: 2167
Credit: 64,403,322
RAC: 5,085
Message 36508 - Posted: 27 Mar 2009, 13:42:13 UTC - in response to Message 36507.  

1. http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=7673807
Current timestep = 49417 of 259248
s/TS = 56.6

2. http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=7731951
Current timestep = 39775 of 259248
s/TS = 3.64

Neither are blue / ice worlds, but the to-completion time is rising, so maybe they\'re turning blue...

#1 looks like it has gone back to the beginning of the model (without resetting the CPU time), so it\'s not an iceworld. Occasionally this happens if Windows is shutdown or rebooted while the model is running. But you won\'t get any more trickles or credits until it reaches the point it last trickled. It\'s up to you whether you want to abort it, or run it to the end.
The workunit #2 is in has other models completed by other Windows/Intel computers, and the s/TS are not rising, so probably isn\'t an iceworld. I\'m not sure why it is not making progress, or not trickling.
ID: 36508 · Report as offensive
RichardRodd

Send message
Joined: 16 Mar 06
Posts: 28
Credit: 3,219,100
RAC: 0
Message 36509 - Posted: 27 Mar 2009, 13:43:32 UTC

Greetings, again.

Further to the post just above, I suspended those two tasks, and the waiting task that started up is behaving the same... a rising to-completion number from the outset.

Can someone help me fathom this out? Might it be because this m/c is running with quite a full hard drive - 10Gb free out of 70.

This is a 2-cpu m/c and BOINC hasn\'t downloaded a second task to me. I found a BOINC discussion thread (http://einstein.phys.uwm.edu/forum_thread.php?id=6389) that suggested BOINC knows this m/c is struggling, but I\'m not sure how to go about trouble-shooting...

Oh, I also upgraded from 6.4.5 to 6.4.7, so it should be a clean installation here.

All help gratefully received - as ever!

Many thanks - Richard
ID: 36509 · Report as offensive
Profile Iain Inglis

Send message
Joined: 9 Jan 07
Posts: 467
Credit: 14,549,176
RAC: 317
Message 36521 - Posted: 28 Mar 2009, 11:01:32 UTC - in response to Message 36490.  

They all appear to be very slow processing, some for up to 3 months. Presumably they are also blue ice-worlds? Your computers have succesfully run other models. I\'d suggest you abort the slow running models. Good luck with the new ones.


They do all display totally Blue on graphics.

While what hagar says is right - there is no point continuing with a slow-processing iceworld - there may be more to it. Quite a few of the models you mention occur in work units where there are Windows/Intel finishers. That suggests to me that there may be machine problems of some sort, rather than an iceworld problem.

Here is your list, linked to the output of a crude iceworld detector ...

7714560
7714126
7703437
7703654
7703204
7703491
7714497
7703217

... there are a lot of singletons and other finishers. One model is a confirmed iceworld, as another model in the same work unit had the same problem (i.e. this one).

You may want to try some of the stability tests in the Read Me, particularly the Prime95 test.
ID: 36521 · Report as offensive
JimMcCarthy_StellarSolns
Avatar

Send message
Joined: 3 Sep 08
Posts: 23
Credit: 41,989,607
RAC: 2,734
Message 36687 - Posted: 11 Apr 2009, 15:06:29 UTC

I\'ve recently had to abort two models (on two different computers) that had become iceworlds:

WorkUnit 6278634, Task ID 8362432.

and

WorkUnit 6277746, Task ID 8354443.

Both machines are Intel CPUs, Windows XP SP3. Temperature graphics on both were blue, with no clouds except around the Antarctic coast. Time to completion was rising instead of falling, and models were no longer uploading trickles.

-- Jim

ID: 36687 · Report as offensive
Profile adrianxw
Avatar

Send message
Joined: 31 Aug 04
Posts: 145
Credit: 2,021,020
RAC: 816
Message 36701 - Posted: 12 Apr 2009, 19:30:45 UTC

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/workunit.php?wuid=6193392
149066 of 259248

Next, not sure, pressed z and cannot see what you want... what it says is...

Atmos model time 13:00
Atmos model date 16:07/2059
Hours elapsed 0418:33:33 (2.26 s/TS) <-- maybe that!
Timestep 149066 of 259248 - Phase 3 of 3
Progress 85.83%

Not blue
Intel Quad under Windows XP
Not over clocked.

Hope that helps, was running on a system not visited everyday so may have been like that for 4-5 days at least.

Best wishes,
Adrian.

Wave upon wave of demented avengers march cheerfully out of obscurity into the dream.
ID: 36701 · Report as offensive
Profile mo.v
Volunteer moderator
Avatar

Send message
Joined: 29 Sep 04
Posts: 2363
Credit: 14,611,758
RAC: 0
Message 36702 - Posted: 12 Apr 2009, 20:10:13 UTC
Last modified: 12 Apr 2009, 20:19:10 UTC

Jim\'s models both sound like typical iceworlds, so well done for noticing and diagnosing them.

Hi Adrian

Here\'s your model\'s workunit. Its sec/ts on its task page shows the first sign of the slowdown and it looks as if Eric\'s model from the same workunit, also with Intel & Windows, has suffered the same fate at the same point. So yes to the iceworld diagnosis.

Re the graphics display, Z eliminates the overlay. 8 toggles the checkpoint countdown on and off. H gives the help menu. You may need to press T to show the temp view which is where the iceworld blue is apparent. P for pressure will also show a different monochrome colour in the case of an iceworld.

Also re diagnosis, the sec/ts displayed in the graphics window is a cumulative average (ie the average since the model started). So in the case of a genuine iceworld that has almost stopped progressing, the sec/ts value shown is not the current very slow speed but it gradually gets worse and worse if the model is allowed to continue.
Cpdn news
ID: 36702 · Report as offensive
Profile Conan
Avatar

Send message
Joined: 6 Jul 06
Posts: 141
Credit: 3,511,752
RAC: 144,072
Message 36827 - Posted: 29 Apr 2009, 12:02:45 UTC - in response to Message 36311.  

I see you've now aborted that iceworld, Mican. Its whole workunit #6278040 is rather interesting. A model on a machine with Linux crashed at about the same point. The model Hagar mentioned that seems to have speeded up was on a Mac. It crashed.

Conan has a model from the same workunit running on Linux and Verstapp has one on Intel/Windows, both further behind. Both Conan and Verstapp are contactable. I'll ask them to watch their models to see what happens well into Phase 4. It will be particularly interesting to see whether Conan gets an iceworld on Linux.


G'Day mo.v,
Just an update, I took a quick look at Verstapp's work unit as it has reached phase 4 and I am still at phase 1. The last TS that he uploaded has just taken a large drop from a consistent 1.2 s/TS to 1.33 s/TS so perhaps he has also gone into an Ice world ?


G'Day again mo.v,

Well it took a bit of time but it looks like Linux triumphs this time.
I have finished the model to completion (took 547 hours).
Time steps stayed the same right through and it shows 'Success', I just have to wait for the credits to be granted.

Here is the WU 8357087
if you want to check it out.
I can't read the data very well so really I have no idea if this was getting close to an iceworld or not, I don't have graphics on Linux enabled.
ID: 36827 · Report as offensive
old_user17401

Send message
Joined: 13 Sep 04
Posts: 60
Credit: 1,124,063
RAC: 0
Message 36850 - Posted: 4 May 2009, 9:32:59 UTC
Last modified: 4 May 2009, 10:06:26 UTC

My first iceworld :) The globe is blue.

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=8423384.

I'm at TS 253706 (99.47%), doing 1.23s/TS, though in reality for well over 24 hours now it has said 2 hours remaining. I'm using Windows XP SP2, C2D E6750, 2.66GHz, with 2GB of RAM. I don't have any higher-priority processes running; the work unit is taking 50% CPU i.e. one saturated core.

Should I let it run (it'll take weeks) or abort? The science results matter more to me than any credits. I notice another user with the same work unit has already aborted at the same point.
ID: 36850 · Report as offensive
Les Bayliss
Volunteer moderator

Send message
Joined: 5 Sep 04
Posts: 7629
Credit: 24,240,330
RAC: 0
Message 36851 - Posted: 4 May 2009, 10:21:41 UTC

I've had several, and I don't mess around with them. If they go blue, I abort them and get another one.


Backups: Here
ID: 36851 · Report as offensive
Profile Conan
Avatar

Send message
Joined: 6 Jul 06
Posts: 141
Credit: 3,511,752
RAC: 144,072
Message 36852 - Posted: 4 May 2009, 11:03:03 UTC - in response to Message 36851.  
Last modified: 4 May 2009, 11:04:23 UTC

I've had several, and I don't mess around with them. If they go blue, I abort them and get another one.


G'Day Les,

A question if you will,
Why has my Work Unit (WU 8357087) which ran on Linux and completed successfully yet every Windows computer that ran this WU all failed and became Ice Worlds.????
ID: 36852 · Report as offensive
Profile mo.v
Volunteer moderator
Avatar

Send message
Joined: 29 Sep 04
Posts: 2363
Credit: 14,611,758
RAC: 0
Message 36853 - Posted: 4 May 2009, 11:15:40 UTC

In the few cases where the cause is computer instability you could correct the computer problem, restore a backup and expect the model to complete normally.

In almost all cases the problem lies within the model itself on an Intel/Windows setup. Nothing would make these iceworlds become normal again. The moment the iceworld develops the model stops processing its data properly and there's no point in trying to continue.

If another model from that WU running on Intel/Windows has developed an iceworld at the same point you know the fault lies in the model, not in your computer.
Cpdn news
ID: 36853 · Report as offensive
Profile mo.v
Volunteer moderator
Avatar

Send message
Joined: 29 Sep 04
Posts: 2363
Credit: 14,611,758
RAC: 0
Message 36888 - Posted: 8 May 2009, 13:19:19 UTC

My MH task 8424391 has become an iceworld in phase 4. Typical monochrome graphics, but it's atypical in that it's on my AMD and has speeded up by about 30%. Ananas has completed a model from the same WU on Intel/Windows.

I'm not happy about this. I'll let the model run in its iceworld condition to complete in about 2 days' time to can see what graphs it produces. I may then restore a week-old backup and run it long enough to see whether it gets past the iceworld moment, which I haven't identified exactly, on a rerun.


Cpdn news
ID: 36888 · Report as offensive
old_user539218

Send message
Joined: 29 Sep 08
Posts: 5
Credit: 4,330,352
RAC: 0
Message 38341 - Posted: 21 Nov 2009, 11:41:09 UTC

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=10450514
172,832
5.0911
Yes
Intel, Windows
No

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=9673466
140,426
4.3909
Yes
Intel, Windows
Yes

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=9629486
97,218
5.7759
Yes
Intel, Windows
Yes

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=9819246
248,446
4.0848
Yes
Intel, Windows
No
ID: 38341 · Report as offensive
Dave Worrall

Send message
Joined: 27 Jan 05
Posts: 16
Credit: 790,158
RAC: 0
Message 38510 - Posted: 11 Dec 2009, 17:50:38 UTC

Ice World...

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=10377841

Timestep 149706 of 259248 - Phase 3 of 3

1.27 s/TS

The globe graphic is blue. The graphic isn\'t stable though, it flashes triangles and other stuff on the globe. This happens all the time if I display graphics, other units have completed ok. Also won\'t close cleanly. Anything to do with dual processors?

Dell Vostro 220, Intel(R) Corel(TM)2 Duo CPU E7300 @ 2.666GHz

Microsoft Windows XP Pro Version 2002 Service Pack 3.
ID: 38510 · Report as offensive
wateroakley

Send message
Joined: 6 Aug 04
Posts: 185
Credit: 27,083,655
RAC: 6,161
Message 38511 - Posted: 11 Dec 2009, 18:15:58 UTC - in response to Message 38510.  

Ice World...
Anything to do with dual processors?
No, not dual-processor related, it sounds like a slow-processing Intel/Windows iceworld. Iain Inglis is collecting iceworld data iceworld appeal if you have a backup. Otherwise abort it and get another one. Good luck.

ID: 38511 · Report as offensive
old_user26115
Avatar

Send message
Joined: 21 Oct 04
Posts: 24
Credit: 207,633
RAC: 0
Message 38543 - Posted: 17 Dec 2009, 10:12:56 UTC

Bad luck, I got an Iceworld after 1 Phase was completed:

A link to the model/ResultID webpage
- ResultID=10528029

A current timestep of that model (on the globe graphic)
- 9799 Phase 2 of 4 (9800 TS in 10 hour ! after finish Phase 1)

The s/TS value (on the globe graphic.)
- 1.34 s/TS , before it was 1.31

Whether the temperature display of the globe graphic is blue.
- yes , blue -42°C

What your processor/CPU and Operating System is (i.e. Intel or AMD on Windows or Linux)
- GenuineIntel Pentium(R) Dual-Core CPU T4300 @ 2.10GHz [Intel64 Family 6 Model 23 Stepping 10]
- Microsoft Windows 7 Home Premium x64 Edition, (06.01.7600.00)

Whether you are overclocking.
- no

I will abort this WU

greetz littleBouncer

ID: 38543 · Report as offensive
Previous · 1 . . . 10 · 11 · 12 · 13 · 14 · 15 · Next

Message boards : Number crunching : Iceworlds & Slowdowns hadsm3/mh - Closed - Discussion

©2024 climateprediction.net