30/10/05 13:21:14|climateprediction.net|Message from server: No work sent
30/10/05 13:21:14|climateprediction.net|Message from server: (won\'t finish in time) Computer on 0.0% of time, BOINC on 99.1% of that, this project gets 33.3% of that
30/10/05 13:21:15|climateprediction.net|Deferring communication with project for 1 days, 23 hours, 59 minutes, and 58 seconds

My computer is on 100% of the time unless there is a power failure and BOINC runs 50% of the time (Scheduled tasks start each program on alternate days{SETI@home cmd line and BOINC} every 48 hours and shuts them down after running 23h 45m), SETI@home Classic (cmd line version) runs the other 50% of the time. I also run BOINC Einstein@home and BOINC SETI@home. Both BOINC and SETI@home Classic are set to run always. Each BOINC project gets 1/3 of resources. So, why the above error?

---

It\'s possible that your computer is overcommitted with the other projects.
The newer versions of BOINC won\'t give your computer new work if it looks like not being able to meet the deadline for the project.
Sulphur is on a tight deadline because the results are needed in about 3 months.
(Although models finished after this will still be accepted.)

Also, you have had a lot of error 5 failures. This may indicate a computer problem.
The most usual reasons are overclocking, and overheating. The latter can be due to numerous reasons: dust on the heatsink, inadequate air flow, high room temperature, etc.
Also, a lot of off the shelf computers have a barely adequate power supply, and the voltage drops too much when lots of programs run at once. Or just cpdn by itself, as it\'s very processor intensive.
When you have a combination of these, you get problems.

There are references to programs that can help find problems if you\'re interested, and Mother Board Monitor can keep an eye on your cpu temperature.

Actually it looks like your time stats got corrupted somehow. They should fix themselves eventually.

If you are impatient you can edit the client_stat.xml file to put more appropriate values in. They should be between 0.0 and 1.0. Make sure you exit BOINC before editing.

---

BOINC WIKI

BOINCing since 2002/12/8

Sulphur is on a tight deadline because the results are needed in about 3 months.

I recently posted with regard to the sulphur deadlines. It would appear that one of my PCs (a 750 mhz pentium) hasn\'t got a snowball in hell\'s chance of completing the result in the available time.

BOINC runs benchmarks. Surely it\'s possible for CPDN to estimate how long a result will take to complete (from the benchmarks), and not send out a result to a PC which can\'t possibly complete it in the time available. OK I will get the credit for the trickles up to the deadline, but then what? If I get zero credit after the deadline, I\'m just going to reset the project, so they won\'t get a completed result. (Human nature.)

Are they still issuing hadsm3, or is it just sulphur at the moment?

Brian

---

Brian

Brian
I answered some of this in your other post.
cpdn doesn\'t use the BOINC bench marks, as they are not very accurate on a project with such a long completion time.
However the programmers have tried to set up the server software so that longer, slower sulphur models are given to fast computers which have allocated a large part of their processing time to cpdn, and slab models to the rest.
This hasn\'t worked too well yet. And multi-project people are continually changing the percentage times in mid-model.

If you start a sulphur model, and it doesn\'t crash or get aborted by the user, you will continue to get credit. And the result will be accepted.

Slab models will continue to be available, as this is a requirement of the Open University\'s climate modelling course.