On Tue, 07 May 2013 21:30:10 -0400, Nicolás Alvarez <[email protected]>
wrote:
2013/5/7, Charles Elliott <[email protected]>:
There are three facts of which you are obviously unaware:
1. It is intensely humiliating to experience the loss of thousands of
workunits, especially after one has spent hours trying to avoid that exact
situation, and after taking exactly the same actions that had avoided that
loss previously.
"Task no longer usable" means the tasks were cancelled on the server.
Maybe whatever test S@H Beta was doing with those tasks was finished,
so they aborted the remaining tasks for that test. If the client kept
processing them, the server would reject the result because the task
was cancelled; that would be a huge waste of time and resources.
I've trimmed the rest to note a side effect based on just that part. The
cancelled tasks and the server aborts which ensued resulted in most hosts
having their quotas reduced to 1 as if there had been a compute error on each
of those aborted tasks.
At one time there was special handling for server aborts which avoided that
effect, but it has since gone missing. I think some thought should be put into
the purpose of reducing quota and which kinds of errors ought to be exempted.
There have been other situations where a server problem has resulted in
transfer errors, for instance, though it would be more difficult to distinguish
those from a client side problem.
--
Joe
_______________________________________________
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.
