If it's a per-project question, then each project is going to have to develop their own solution. The best BOINC can do is provide a few hooks.
... and if it has to be done per-project, then there isn't a whole lot to discuss here. The point I was trying to make is: what problem is this discussion really trying to solve? If the problem is to instantly catch a bad host, and cut it off, I don't see a way to do it. If we can let go of the word "instantly" then I see a way for BOINC to stop the longer term damage. On 5/28/2010 4:53 AM, Richard Haselgrove wrote: > That's a question for each project, not BOINC centrally. > > It comes in two parts: > > Does it need fast (valid results)? > (e.g. to generate dependent work, or to purge processed files from a > small server) > > Can it cope with fast errors? > (does the demand for new work from a fast-erroring host cause server > problems) > > >> Seems to me that the really fast methods of catching a broken host >> aren't very good, and the good methods aren't very fast. >> >> Does BOINC need fast? >> _______________________________________________ boinc_dev mailing list [email protected] http://lists.ssl.berkeley.edu/mailman/listinfo/boinc_dev To unsubscribe, visit the above URL and (near bottom of page) enter your email address.
