This is essentially the issue quota system that MW uses at the moment. You have an issue quota of n tasks; MW used CPU cores times 6, which is farbled if you have and only use a GPU, and this would allow this limit to be adjusted (if desired) ...
In this case they could set the quota at 48 for each (app version, host) pair (or whatever) and retain the daily limit of 10,000 currently set per system. Should this be a "hard" number that has to be configured for each (app version), or will there be a mechanism so that it is self-adjusting (if allowed by the project)? I mean, MW will want a hard number because of their server issues, SaH likely would want a number that can be automatically adjusted... Though that drives the question, what about Anon Platform... Still, I can hear the howls from SaH already, even if this won't really change anything ... :) On May 24, 2010, at 1:07 PM, David Anderson wrote: > Here's an idea: the server keeps track of the # of jobs sent today > for each (app version, host) pair. > > Tweak: when validate a job, decrement this count. > > That way, as long as a host returns correct results, it will keep getting > results even if its daily quota is low. > > Lynn W. Taylor wrote: >> My first thought is that there needs to be some sort of initial quota >> that should probably be lower than 100 -- someone might try BOINC but >> uninstall right away, or not like how it makes their computer run 100% >> or who knows what -- and uninstalling shouldn't abandon a bunch of work. > _______________________________________________ > 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. _______________________________________________ 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.
