> [email protected] wrote:
>> Paul:
>>
>> The events that drive rr_sim:
>
> OK, so trying a little KISS:
>
>> RPC complete:  We have now committed to some more work.  Is there 
>> anything
>> that needs to start right now to complete on time.
>
> Nope. No need. The request should be reasonable enough and the WUs
> committed to should be reasonable enough that there should be no change
> to the presently running work.
>
>> File Download Complete:  We now have a task ready to run.  Does it need 
>> to
>> get started right now?
>
> Again, should be no change for the above reasoning.

Sorry, slight disagreement with Martin here.

1) RPC complete.
Tasks are quantised. The client can request 1 second, but it will receive 
more. There are reported cases (rare, and as yet not explained) when a 
perfectly reasonable small request is responded to with a great deluge of 
work (http://setiathome.berkeley.edu/forum_thread.php?id=51410#853921). One 
extra test iteration - perhaps skipped if no work is allocated - seems a 
small price to pay for a pair of braces (suspenders) to reinforce the belt 
supplied by the server.

2) Download complete
As somebody pointed out a long time ago in this thread, owing to bandwidth 
congestion or whatever, a file download may complete many hours or even days 
after it was initiated - and it may need to run immediately. We may know 
that already if one of the other triggers fired, but we may not - and in 
either event, we need to trigger at least an 'enforce' test post-download. 


_______________________________________________
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.

Reply via email to