There are a few more places where the test is needed.
When a scheduler RPC is complete. The RPC can have committed the computer
to much more work than it already has. It can also have removed some work,
including work that is currently running.
When work errors out, as opposed to successful completion (these are
different places in the code).
The tests are already done only when it is needed. If there is a problem
with the test, we need to fix the test, not the frequency. Fixing the
frequency will not help if the test is wrong.
jm7
TarotApprentice
<tarotapprentice@
yahoo.com> To
Sent by: BOINC dev
boinc_dev-bounces <[email protected]>
@ssl.berkeley.edu cc
Subject
04/25/2009 08:30 Re: [boinc_dev] 6.6.20 and work
PM scheduling
> Date: Sat, 25 Apr 2009 14:31:01 -0500
> From: "John Keck" <[email protected]>
> Subject: Re: [boinc_dev] 6.6.20 and work scheduling
> To: "Paul D. Buck" <[email protected]>, "Richard Haselgrove"
<[email protected]>
> Cc: BOINC Developers Mailing List <[email protected]>,
John McLeod <[email protected]>
> Message-ID: <49ac83b79e5d462cbe9dc89290e34...@tsi5200w>
> Content-Type: text/plain; format=flowed; charset="iso-8859-1";
reply-type=original
>
> I definitely agree with Paul; I don't see a good reason to test for a CPU
> reschedule more than once a minute and 5 to 10 minutes seems more
reasonable
> with the exception of a resource going idle. In the best case we are
wasting
> cycles that could be used by the science app. In the worst case we kill
> tasks that have trouble dealing with multiple stops and restarts.
>
> john
I'd suggest:
a) Download finishes (more work has arrived - we need to check deadlines)
b) Task completes (we have a free resource)
or if none of the above every 5 minutes
MarkJ
The new Internet Explorer 8 optimised for Yahoo!7: Faster, Safer,
Easier.
_______________________________________________
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.
