Message boards :
Number crunching :
Let me understand how credit works
Message board moderation
Author | Message |
---|---|
Elawise Send message Joined: 19 Jul 99 Posts: 12 Credit: 52,232 RAC: 0 |
I have 6 machine running boinc 4 running 24/7. I have not recieved ANY credit for several days. In reading these boards I understand that 3 WU's are sent out and no credit is given until all three are returned and validated. If one does not report correctly then no one recieved credit and those three are sent out again? It one is not sent back by the deadline then no one recieves credit? Pardon me but how lame is that. I realized that proper vailidation requires the redundance. But where is the redundacy for us? If three valid results are required that why only send three. Shouldn't the WU's be sent to 5 or 6 clients to ensure they are returned on time and with correct results? That way if there were any invalid results they could be tossed imediatly without haveing to reschedule a download of the WU to three new machines. I would think it would be less stress on the severs to send 5 out, and validate those and as long as the majority agree (returned correctly on time with agreeing results) then the server would not have to re-schedule the WU etc. But hey, I'm only an engineer...... |
Pascal, K G Send message Joined: 3 Apr 99 Posts: 2343 Credit: 150,491 RAC: 0 |
The bird says Links to Documents by Project or How I found out all the Stuff I Needed to Know About BOINC My name is Pascal and this message has my approval... It is 10 oclock, do you know what your WUWUs are doing tonight... |
Paul D. Buck Send message Joined: 19 Jul 00 Posts: 3898 Credit: 1,158,042 RAC: 0 |
> It would be a waste of CPU time to send the WU to 5 or 6 hosts initially if > only 2 or 3 results are really needed. Most of the time, the three hosts > report back valid results by the deadlines. So sending the unit 6 times would > waste the processing power on 3 hosts, add twice as many database records to > the already heavily loaded database server, and require the already heavily > loaded validator to look at twice as many results. Good points all, but the most important one is that right now Classic is "wasting" the available CPU cycles. Instead of re-hashing something that has little to no importance is wasteful. Now, with multiple projects in the BOINC Family we can spread those cycles around to do work that is going to count as in making a real contribution to science and not just an increase in the participant's credit total ... <p> For BOINC Documentaion: Click Me! |
©2024 University of California
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.