I\'m not sure if my models running properly. Have trickles been uploaded? I tried setting the proxy server as suggested and it seemed to block internet access. I\'ve finished a Hadam and Hadsm model and I\'m not sure if they completed properly. I can\'t find any previous models to view in CPview.

---

Hello Jo - and welcome to the message board.

There aren\'t any completed models for your active computer, 846028, but the two running models trickled on the 17th May. Was the proxy changed after that?

The PC looks to be a laptop and some care has to be taken with that kind of machine in case it shuts down in a power-save mode - which will very likely crash BOINC. It\'s a good precaution always to exit BOINC manually before closing down.

Iain

I changed the proxy today but I\'ve changed it back again. If the models trickled on the 17th, does that mean they\'re okay at the moment? My models say they\'re restarting tasks again, is that normal and is that a trickle point. I can\'t remember any trickle messages from 17th but those messages have now disappeared.
I\'ve set my laptop to hibernate when I shut the lid, should I exit Boinc before hibernating. After reading other threads I have started being more careful; pausing Boinc for updates etc.

19/05/2008 16:01:47||Starting BOINC client version 5.10.13 for windows_intelx86
19/05/2008 16:01:47||log flags: task, file_xfer, sched_ops
19/05/2008 16:01:47||Libraries: libcurl/7.16.1 OpenSSL/0.9.8e zlib/1.2.3
19/05/2008 16:01:47||Data directory: C:\\Program Files\\BOINC
19/05/2008 16:01:48||Processor: 2 GenuineIntel Intel(R) Core(TM)2 CPU T7200 @ 2.00GHz [x86 Family 6 Model 15 Stepping 6]
19/05/2008 16:01:48||Processor features: fpu tsc pae nx sse sse2 mmx
19/05/2008 16:01:48||Memory: 2.00 GB physical, 3.85 GB virtual
19/05/2008 16:01:48||Disk: 142.31 GB total, 118.48 GB free
19/05/2008 16:01:48|climateprediction.net|URL: http://climateprediction.net/; Computer ID: 846028; location: (none); project prefs: default
19/05/2008 16:01:48||No general preferences found - using BOINC defaults
19/05/2008 16:01:48||Reading preferences override file
19/05/2008 16:01:48||Preferences limit memory usage when active to 1023.18MB
19/05/2008 16:01:48||Preferences limit memory usage when idle to 1023.18MB
19/05/2008 16:01:48||Preferences limit disk usage to 9.31GB
19/05/2008 16:01:48|climateprediction.net|Restarting task hadcm3istd_0knt_1920_160_15987196_7 using hadcm3i version 544
19/05/2008 16:01:48|climateprediction.net|Restarting task hadcm3istd_0knv_1920_160_15987198_1 using hadcm3i version 544
Jo

---

1. The \'restarting\' message appears each time BOINC starts: it just means that the tasks are starting again from where they left off. It doesn\'t mean that a new task is starting.

2. Definitely exit BOINC before hibernating. Since it\'s pretty much impossible to remember every time, it might be a good idea to make a backup from time to time. If a model crashes then it can be restored and will carry on from where it was backed up. Instructions are here.

If you let it run for a bit then we\'ll see if any trickles appear. There\'ll be about one per day of run time - longer if the machine is turned off. For the models you\'re running - \'coupled\' models - a trickle is sent to the server every model year.

Okay; that should mean I next see a trickle when 1.25% is completed. I\'ve ordered an external drive to backup onto. What about the messages; do they disappear after a couple of days normally?
Jo

Everything in the Messages tab of BOINC Manager is cleared out every time BOINC Manager starts, though it is kept somewhere on disk.

The first trickle will be at half of 1.25% (i.e. 0.625%) because your two are 160-year models (the recent batch of 80-year models would indeed be 1.25%).

Messages are stored in the file stdoutdae.txt in the directory you installed BOINC to. The file has a 2MB size limit, which causes the current file to be renamed to stdoutdae.old and a new stdoutdae.txt file to be started.

---

"The ultimate test of a moral society is the kind of world that it leaves to its children." - Dietrich Bonhoeffer

Everything in the Messages tab of BOINC Manager is cleared out every time BOINC Manager starts, though it is kept somewhere on disk.

The first trickle will be at half of 1.25% (i.e. 0.625%) because your two are 160-year models (the recent batch of 80-year models would indeed be 1.25%).

Hi, I realize that this is a bit off topic, but, I was wondering if there is any way that I can ensure that net time I get one of the 80 year coupled models instead of the 160 yrs models. I like to run coupled models because of the increase scientific value of using a realistic multi-level ocean, but, on my machines the 160 yr models are a very long haul. Each takes about 7 or 8 months to complete running pretty much 24/7.


Jim

---

Sorry, no.
At least, not the way things are organised at present.
It depends on what is in the queue for the coupled models.

The slab models have scientific value for a project in Germany, which has requested them. And there is another version, not yet released, which has 4 phases, to test some ideas.
Apparently researchers around the planet can see interesting possibiities in variations of the slab models, and more sub-projects using them may show up over the coming years.

Ive been running an HadCM3 160yr model for 28hrs now and it hasnt attempted to trickle yet, its currently at 1.359% (56400 steps). Is there a command to force a trickle so I can see any error message?

adleberg

No. There\'s no command to do this.
The only way to get to a trickle is to be patient.

I\'ve forgotten the % point at which it reaches a trickle, but it will be on the 4th of December.

Use the Show graphics button to look at the model\'s display. If you haven\'t done so yet, press Z to remove the left hand overlay, and then 8 to display the countdown timer.
Among the info now visable will be the model\'s day/month/year.

---

Backups: Here

curiously, ctrl-z actually freezes the graphical display, the model keeps computing but the display is locked..the model is on 13/02/1923, so its passed december three times. Its not that im over concerned about credit, i just dont want to finish 1500 hours and find out there was an error :)

NOT <ctrl> Z, just plain Z.

You might like to have a look through the READMEs linked from my signature. There\'s lots of info there.

As for running for 1500 hours with an error, don\'t worry. If there\'s an error, the model will crash.

Are you sure that there\'s been no trickles? Perhaps they are stuck in the project folder. i.e. Something like: C:\\Program Files\\BOINC\\projects\\climateprediction.net
If so, they will have names starting with trickle_up_hadcm3_

---

Backups: Here

Ive been running an HadCM3 160yr model for 28hrs now and it hasnt attempted to trickle yet, its currently at 1.359% (56400 steps).

HadCM3 models should create a trickle every 0.625% (25,920 timesteps), so you should have uploaded 2 trickles so far.

Checking my own results it looks like the server is still catching up with the trickle processing backlog that built up during the recent database problems. My most recent trickle was returned at 2139 UTC last night and processed at 1126 UTC this morning (i.e. the backlog was just under 14 hours at that point).

The result page for your active model shows one trickle was processed at 1013 UTC. The other one will be in the pending queue.

---

"The ultimate test of a moral society is the kind of world that it leaves to its children." - Dietrich Bonhoeffer

NOT <ctrl> Z, just plain Z.

You might like to have a look through the READMEs linked from my signature. There\'s lots of info there.

As for running for 1500 hours with an error, don\'t worry. If there\'s an error, the model will crash.

Are you sure that there\'s been no trickles? Perhaps they are stuck in the project folder. i.e. Something like: C:\\Program Files\\BOINC\\projects\\climateprediction.net
If so, they will have names starting with trickle_up_hadcm3_

Hi together,

i got ecactly the same problem, the project reached the 4th of July 1923 and has not sent any trickles so far...

Cheers

---

Hi together,

i got ecactly the same problem, the project reached the 4th of July 1923 and has not sent any trickles so far...

Cheers

Three models have been downloaded to your quad:

1. 7948102 (2 trickles)
2. 7511012 (no trickles)
3. 7509907 (1 trickle)

Best to let #2 run for a bit, to see whether it has a problem or not. It looks OK at the moment.

I am running two Slab models. One is progressing fine and has nearly finished. The other, which was started the same day, has started running backwards!! The last trickle was sent on 12 July, and in the last few days the time to completion has been going UP and the progress has been stationery - or actually going DOWN! The task ID is 7418486.

Two possibilities:
1) It\'s turned into an \"ice world\" Info
2) You interrupted the model at the phase change and before it reached the 1st savepoint in the next phase. (No real help thread, just replies in various places, and warnings in News and Announcements.

My guess is 2) (No more trickles will be registered until the model catches up to where it was.)

---

Backups: Here

If the globe graphics look normal it will have gone back to repeat phase 1 & 2 or just phase 2. If the graphics are seriously abnormal it will have become a slow-processing iceworld.

---

Cpdn news

Two possibilities:
1) It\'s turned into an \"ice world\" Info
2) You interrupted the model at the phase change and before it reached the 1st savepoint in the next phase. (No real help thread, just replies in various places, and warnings in News and Announcements.

My guess is 2) (No more trickles will be registered until the model catches up to where it was.)

The graphics certainly look like an ice world! I\'ve taken a screenshot but can\'t see a way of attaching it. Details are:
Model date: 10/06/2060
hours elapsed: 505:03:50 2.66s/TS
timestep: 164597 out of 259248
Progress: 87.83%

I\'ve looked at the workunit (615042) and notice that two other tasks (7418482 and 7418481) seem to have stalled at the same point.

The last backup I have is dated 15 June but this includes another model running normally and 95% complete.