Since 1FEB05, my laptop has been running and failing repeatedly with "Client Error".

So far I have Results = 16. Of these, 14 have resulted in "Client Error", one is in process Outcome="Unknown", and one seems to have become lost, also Outcome="Unknown".

Most recently received workunit was accompanied by the message
"climateprediction.net - 2005-04-11 05:47:24 - Starting result 2f91_300134223_0 using hadsm3 version 4.12". Yes, it has just started a new work unit yet again today.

The system this is running on is a Dell Latitude D400 laptop.

CPU Type = GenuineIntel Intel(R) Pentium(R) M processor 1300MHz

Operating System = Microsoft Windows 2000 Professional Edition, Service Pack 4, (05.00.2195.00)

So far it has been granted Total Credit = 9073.68, but is this truly useful, since no work unit ever completes?

If the computer is merely generating "credits", but contributing nothing to science, I think it is best to deinstal.

I have four other computers running this project. One of them has had 2 "Client Errors". They have identical "received" time stamps (it is a dual processor system, so there are two instances active).

The other three computers are still on their first work unit.

Insight anyone?

Thanks

John

---

Heat is your enemy. Anything you can do to improve ventilation around/through your laptop would lessen the chances of heat-related failures. Tilting up the back end of your laptop could help, while ensuring none of the vents are blocked. If the laptop has been around awhile, compressed air may clean dust out from around the processor area. For a more expensive solution, there are USB powered cooling pads for laptops that can help keep it cool.

Some laptops can be made stable through better cooling, and others may not. If the above tips don't help with stability, it may be best to take that PC off the project as you are correct in that quickly aborted results don't really help the project investigators out in any way.

> Heat is your enemy. Anything you can do to improve ventilation around/through
> your laptop would lessen the chances of heat-related failures. Tilting up the
> back end of your laptop could help, while ensuring none of the vents are
> blocked. If the laptop has been around awhile, compressed air may clean dust
> out from around the processor area. For a more expensive solution, there are
> USB powered cooling pads for laptops that can help keep it cool.
>
> Some laptops can be made stable through better cooling, and others may not.
> If the above tips don't help with stability, it may be best to take that PC
> off the project as you are correct in that quickly aborted results don't
> really help the project investigators out in any way.
>

GEOPHI and JOHNFKITCHEN:

Right on, geophi. In other threads you will see that I have had problems of aborting or at least losing models prematurely.
I have installed a program recommended on another board. It is HDD Health, v2.1 Beta, by panterasoft (www.panterasoft.com) .
In today's readings by HDD Health, it indicated a beginning temperature "value" (ID C2) of 230 under the S.M.A.R.T. readings, and this has reduced to 177 in about 30 minutes of operation after a complete shutdown over the weekend.
The values may not be actual temperature. I am not totally familiar with the workings of HDD Health yet. BUT, this is the only reading which is highlighted as abnormal.
I am taking further measures to avoid temp rise, and will restart BOINC models when I am certain that temp is not interfering.

Thanks geophi, and good luck John

---

Thank you John Eric Hopkinson and Geophi,
This makes sense to me. The laptop does get very hot when the CPU is working hard, even though I DO keep it tilted up to improve airflow. I have seen a number of error messages saying that "memory could not be read". This may be a hardware error caused by heat, or it may be software error according to some sources.

Anyway, doing something about heat seems like a good plan.

Thanks!

John

---