Posts by ChertseyAl

41) Message boards : Number crunching : Random (and low) credit (Message 53)
Posted 23 Aug 2011 by ChertseyAl
Post:
If anyone has any ideas as to what the problem could be, I am open to suggestions.


No idea here. These three tasks all from the same host, all run in sequence, all the same app version got granted wildly different credit. The lowest one was about right, the next was too much, and the highest was *way* too much (not complaining, makes up for the really mean ones!):

16463 15670 21 Aug 2011 18:03:55 UTC 23 Aug 2011 5:44:40 UTC Completed and validated 0.00 42,772.81 208.63 Get Decics with Bounded Discriminant v1.02
16462 15669 21 Aug 2011 18:03:55 UTC 23 Aug 2011 11:12:46 UTC Completed and validated 0.00 46,902.24 132.07 Get Decics with Bounded Discriminant v1.02
16461 15668 21 Aug 2011 18:03:55 UTC 22 Aug 2011 18:20:53 UTC Completed and validated 0.00 49,958.61 444.11 Get Decics with Bounded Discriminant v1.02

My only guess is that it's something to do with the widely discussed 'CreditNew' system, but I'll leave that for other to discuss :)

Al.
42) Message boards : Number crunching : Random (and low) credit (Message 43)
Posted 22 Aug 2011 by ChertseyAl
Post:
And then, just to prove me a liar, the last 4 WUs I returned had sensible credit ;)

Looks like the standard BOINC granting method now.

Perhaps it was the truly magically 'CreditNew' getting it's act together ;)

Al.
43) Message boards : Number crunching : Random (and low) credit (Message 37)
Posted 22 Aug 2011 by ChertseyAl
Post:
I've only crunched a dozen or so WUs so far, and I can't figure out the credit granting scheme.

Looks like you might have changed to a fixed credit scheme as 27.35 seems a common value, but thats awarded for WUs ranging from 19,744 seconds to 67,079 seconds. But then there are some other random amounts of credit for shorter WUs.

My daily RAC varies from 70 to 307 across similar spec hosts, with an average of 144. For these hosts I get an average of 381 across all the other projects I crunch (all hosts are P4HT, hence the low figure).

Looks like the credit needs at least doubling to be even remotely 'fair' ;)

Al.
44) Message boards : Number crunching : Doesn't run on XP SP1? (Message 32)
Posted 21 Aug 2011 by ChertseyAl
Post:
This WU ...

http://stat.la.asu.edu/NumberFields/result.php?resultid=11533

... died instantly with:

<core_client_version>5.10.45</core_client_version>
<![CDATA[
<message>
There are no child processes to wait for. (0x80) - exit code 128 (0x80)
</message>
]]>


I'm assuming it's because that particular XP host is only at SP1?

Not a problem, just thought I'd mention it :)

Al.
45) Message boards : News : Only Linux and Windows currently supported (Message 18)
Posted 19 Aug 2011 by ChertseyAl
Post:
Great! New project, and it's a mathematics one too :) Count me in!

Al.


Previous 20


Main page · Your account · Message boards


Copyright © 2024 Arizona State University