I have now had at least 72 days over which the user/host average has remained at 482.

As far as I can tell, units are generally getting downloaded and results uploaded OK. Credit claimed and granted has been 0 since 18 March 2006. The 482 was credited in one go on 16 February 2006.

Is there something I don\'t understand about the way this works?

Thanks for any advice.

---

Looking at your host results, there are huge number of model\'s crashes.

is your machine overclocked?
overheating?
what antivirus solution?

I would first suggeste to deal with problem of machine stability and/or crashing models.
It would be also good to upgrade to latest BOINC version - 5.4.11

Credit is based on trickles uploaded. Some models were not running long enough to rech and upload upload as least one trickle...which is why you got zero credit for such cases.
But the are obviously models that lasted long enough to reach some trickles, which you should have been credit for, I think...

---

<i>phpBB forum for CPDN, all are </i><a href="http://www.climateprediction.net/board">invited</a>

Thanks for the reply.

The machine is not overclocked (2.8 Mhz Pentium, 1 Gb RAM, 80 Mb drive with 37 Gb free).

I use Norton AV which is kept up-to-date every day.

You say there are a huge number of crashes, but most of the recent results report \"Success\" - which I assumed meant they were being completed OK. Am I wrong?

I have had problems before with the model, when it was crashing regularly, but one of the Mods suggested some ideas which allowed it to continue. However, I am not getting credit for it. I\'m not that bothered about the credit - although it would be nice to have the acknowledgement! - but I don\'t know if the time my PC is putting in is useful. If it is, fair enough; if not, I might as well link to another project.

---

Each of the models take a minimum of 3 months on a very fast computer, running continuously, to complete.
There is a problem with the success message, possibly caused when the model fails in such a way that BOINC thinks that it\'s finished, displays progress as 100%, and uploads this to the server. Or perhaps it\'s a problem at the server end.

A completed model has a cpu time of over 8 million seconds, and over 4 million timsteps.

If you click on the number under \"Result ID\" for any of your models, you can see the messages about why each failed.

Norton is one of several AVs notorious for causing failures in models.
It\'s advisable to exclude the entire BOINC folder from both scheduled and manual virus scans.

In the case of Norton, it locks each file before scanning it, and if this lock period co-incides with when the model needs it, then the model crashes.

You are getting the dreaded -161 error, take a look here http://climateapps2.oucs.ox.ac.uk/cpdnboinc/forum_thread.php?id=4231

---

BOINC Wiki

OK, I have excluded the BOINC folder from AV scans. If it doesn\'t seem to have fixed it in a few weeks I will just get rid of the climate prediction and/or BOINC. I don\'t understand much of this, and it seems as though it\'s a waste of my PC\'s time which I could use on something else. I had problems with this software previously, and I can\'t be bothered if it can\'t be fixed reasonably quickly. It\'s a shame, because I do think we need to do whatever is possible about global warming, but...

---

Model crashes are a lot less of a problem if you\'re lucky enough to have a backup - here\'s a simple way of doing them (I do it weekly).

* Right-click on the icon, \'exit\'
* Navigate to \'c:\\program files\\\' with My Computer or Windows Explorer
* Right-click on \\climate change experiment\', copy, paste
* (if you get a \'files in use\' error, reboot, and repeat the above steps)
* Now you can resume the model by running the manager again.

---

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