Given that Moore's Law applies to server hardware too, do we have any evidence 
that any current BOINC project server would be over-stressed by one RPC per 
hour per active host? CPDN have adopted a 3600 second backoff between RPCs, but 
I believe this is mostly to prevent individual hosts downloading too many new 
tasks when available.

I've been running v7.2.36 with the one-hour reporting for three days now, and 
I'm pleased with it: with a fast GPU at SETI, I'm reporting up to 20 tasks at a 
time, although slower CPU projects can report singly. With piggyback work fetch 
applied to the reporting RPCs when needed, I'm seeing a more even application 
of Resource Share between projects, with less of BOINC v7's tendency to lurch 
between 100% saturation for Project A and 100% for Project B (as anticipated in 
ClientSchedOctTen).



>________________________________
> From: "McLeod, John" <[email protected]>
>To: "McLeod, John" <[email protected]>; David Anderson 
><[email protected]>; "[email protected]" 
><[email protected]> 
>Sent: Monday, 6 January 2014, 16:02
>Subject: Re: [boinc_dev] RFC: Don't let a task be in the state "Ready to 
>report" if its files were uploaded immediately before
> 
>
>Instead of reporting after an hour which will have the consequence of 
>reporting almost every task on its own, how about reporting no later than 48 
>hours prior to deadline (this will cope with one day vacations for most 
>machines with some added slop if there is a delay turning the machine back 
>on), and report no later than min_queue + 24 hours (this will avoid a 
>reporting late because of turning off the machine a few minutes early because 
>of a longer vacation).
>
>-----Original Message-----
>From: boinc_dev [mailto:[email protected]] On Behalf Of 
>McLeod, John
>Sent: Monday, January 06, 2014 10:46 AM
>To: David Anderson; [email protected]
>Subject: Re: [boinc_dev] RFC: Don't let a task be in the state "Ready to 
>report" if its files were uploaded immediately before
>
>Changing it to one hour will essentially disable multi task reporting for most 
>projects.
>
>-----Original Message-----
>From: boinc_dev [mailto:[email protected]] On Behalf Of David 
>Anderson
>Sent: Monday, December 30, 2013 1:31 PM
>To: [email protected]
>Subject: Re: [boinc_dev] RFC: Don't let a task be in the state "Ready to 
>report" if its files were uploaded immediately before
>
>The policy for when to report tasks is in
>CLIENT_STATE::find_project_with_overdue_results().
>One of the clauses is that tasks are reported within 24 hours of completion.
>I changed this to 1 hour.
>-- David
>
>On 27-Dec-2013 10:11 AM, Toralf Förster wrote:
>> -----BEGIN PGP SIGNED MESSAGE-----
>> Hash: SHA256
>>
>> On 12/27/2013 02:30 PM, Toralf Förster wrote:
>>> The topic was discussed kin the past - yes. And I do understand to
>>> objection to not overload the network resources of a BOINC server
>>> with tasks which could be done a day later or so.
>>>
>>> But today I ran (again) into a case, where a task was solved fine
>>> and its files were uploaded successfully to to the World Community
>>> Grid server around 23th of December, but the task itself stayed in
>>> the state "Ready to report". Then the computer was s2disked.
>>>
>>> Today I came back - and the task is marked as "Too late" (the dead
>>> line was yesterday).
>>>
>>> Well, Christmas is a rare case (within a year) - but I think,
>>> there's place for an improvement to upload *and* report a task
>>> immediately instead of reporting it (at least at my system almost)
>>> later.
>>>
>>>
>>>
>> FWIW here're the last 500 lines of /var/lib/boinc/stdoutdae.txt :
>> http://bpaste.net/show/162219/ 
>>
>> and I do refer to this task :
>>
>>
>>     
>> Workunit Status     
>>     
>> Project Name:         FightAIDS@Home - Vina
>> Created:         12/16/2013 11:04:43
>> Name:         FAHV_x1ZTZ_prnotAS2_0621789_0533
>> Minimum Quorum:         1
>> Replication:         2
>>
>>
>> - --
>> MfG/Sincerely
>> Toralf Förster
>> pgp finger print:1A37 6F99 4A9D 026F 13E2 4DCF C4EA CDDE 0076 E94E
>> -----BEGIN PGP SIGNATURE-----
>> Version: GnuPG v2.0.22 (GNU/Linux)
>> Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/ 
>>
>> iF4EAREIAAYFAlK9wrgACgkQxOrN3gB26U5YpwD/eBz0CbBnc2wZMjTS74tFnRFW
>> yVoTLxpMhLacpIy/jcwA/j3CYUDvmjrMuCzKpFTGaJYO5vYjILvVKsEsroAKTf1X
>> =uBvX
>> -----END PGP SIGNATURE-----
>> _______________________________________________
>> 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.
>_______________________________________________
>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.
>
>
_______________________________________________
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