I recently found that running 2 climate predictions on a intel 6600 made the system sluggish , whereas running 2 rosetta\'s was not so bad I am assuming the data buss traffic created by CPDN is the culprit , so I went about setting CPDN up to stop when working but let Rosetta continue to run. well it ain\'t working out so good. either they both run or they both stop . in home prefs I made each one run on one processor but I told CPDN not to run while computer is in use and Rosetta TO RUN while computer is in use, but while in \"run by prefs\" both will run , any ideas?

Hi Robby

After you\'d changed your preferences did you go into boinc manager, highlight that project and click Update to tell the project server about the change?

The easiest thing would probably be for both Rosetta and CPDN to change your preferences for this computer to use one core only. Then only one task would ever run.

Your Intel 6600 dual-core looks similar to mine, though I\'m not sure whether you have one or two of these

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/hosts_user.php?userid=11898

Mine runs two Hadcm coupled models side-by-side all the time without slowing the computer down, and you certainly have more than enough RAM for these models. Are you running the CPDN screensaver? If so, I\'d advise disabling it as it slows down the crunching and can get in the way of ordinary computer use.

You seem to have been getting a lot of models you didn\'t want and had to abort. In the Projects tab of boinc manager, highlight CPDN and click No new tasks. The day you do want a new model you\'ll have to click the button again.

The second Intel 6600 on your list of computers completed two Hadsm slab models successfully but has also had a lot of model crashes with exit code 1. This could possibly be a recurring graphics problem, so make sure the screensaver\'s disabled and it would be a good idea to update the graphics card driver (free download from the web). It could also be that you\'re turning the computer off without suspending the model and exiting from boinc first. Or maybe you\'re stopping the climate model through Task manager.

For advice on running climate models, go through my signature to the README about Crashes and other problems. In that README I recommend item #5 by Mike and #6 by Thyme Lawn.

Hope that helps.

Mo

---

Cpdn news

On-site-pref settings eventually propagate to ALL projects to which you\'re connected.
With BOINC versions 5.10.*, you can set up prefs on your computer which DON\'T propagate.

And these versions of BOINC also set a low default value of ram which BOINC is allowed to use, so you may not have as much available as you think.
There are 2 settings, one for when your computer is in use, and the other for when it\'s idle.
Increasing the values of these may make your computer less sluggish.
(e.g. 100% when idle may allow BOINC to run in memory, rather than use a lot of hard disk swap space, and so it will get-out-of-the-way faster when YOU start to use it.)

---

Backups: Here

You seem to have two E6600s, one has much faster benchmark figures than the other. Are the two PCs the same, or are they different?

(I\'m wondering if there is insufficient CPU cooling on one of the PCs, and it\'s OK on the other).

Which of these is the PC which is affected when CPDN runs on both cores?

---

I'm a volunteer and my views are my own.
News and Announcements and FAQ

Thanks for the replies !!! I did put the prefs to run only one core (CPDN and Rosetta) but on my fast machine ...it has for the third time started a second CPDN
I have updated like crazy , and have caught the servers down plenty . on an interesting note when reading prefs for Rosetta the messages said it was getting them from CPDN ,this was in my messages . if the CPDN errors were in the past week then it may have been from me aborting the second task , also the fast computer is running at 3.4 ghz and is orthos stable running no higher than 132 f presently 124f.
is there a way i can post a screen shot of my settings ( I have never done this )
note one computer was running only CPND and the other Rossetta for simplicity but the CPDN was sluggish and it was being turned off so this was split across 2 computers to try and keep things running.

Just got through going over my prefs AGAIN! and found in CPDN prefs where you tell it what kind of model you prefer and found it was set to school NOT home , this would explain the addition of a second model on the fast machine, I will give the servers about an hour to catch up then I will update again and see if this clears things up Robby

In the Projects tab of your manager, click on climateprediction, and then click the No new tasks button.
This will stop the manager from requesting new work each time you try something.

---

Backups: Here

it looks like the local prefs on each machine is the problem it is either run while in use, or not Robby

I haven\'t had a need to run one computer differently one 2 projects, so I\'m not sure if it can be done , but you should be able to do this if you have a climate model on one computer, and your other projects on your other computers, and then use different venues for the different computers.

e.g. Have the \'climate\' computer set up to use the Home prefs, and the \'others\' computers set up to use the Work prefs. This way you can set them differently.

---

Backups: Here

well guess what !! I went to each computer go to advanced , preferences, then in the upper right, select clear ;;NOW CPDN does NOT run and Rossetta DOES easch one is set to run on one cpu one is home one is school make one run always an one run by prefs do not use the local prefs in the advanced tab in CPDN select the type of models you want to run and school , home ,work thanks to everyone s input and I hope this helps someone Merry Christmas Robby

I actually screwed up checking clear;that made it read the project prefs I just happened to catch 3 systems running rosetta and not CPDN while active ,it had on each computer selected one cpu total !! not per project ,putting it back to 2 always runs two ,back to this ain\'t going to work Robby

This is basically about getting BOINC to do different things on different projects, so perhaps you should post here on the BOINC/dev boards, where the BOINC \'experts\' \'live\'.
The first one, BOINC Manager is probably the most appropriate.

---

Backups: Here