cvs commit: src/usr.sbin/powerd powerd.c

[Hajimu UMEMOTO]

Hi,

>>>>> On Sun, 28 Aug 2005 22:36:38 -0700
>>>>> Nate Lawson <nate at root.org> said:

nate> Hello, I am back from vacation.  As Kevin and Bruno will attest, I was 
nate> not happy to go down this path since you can't make everyone happy 
nate> without a proper predictive algorithm.  For small numbers of levels, 
nate> this algorithm works fine.  For large numbers of levels, it can 
nate> oscillate just as much as the previous algorithm when there is a 
nate> periodic load.  I do not think you should add an option to tune the 
nate> parameters as this algorithm should be removed as soon as we have 
nate> something better.  I don't want to make it permanent by adding 
nate> user-visible flags.

Okay, it sounds reasonable to me.  I'll go along with your plan.

nate> The right fix is the project that was started by a Summer of Code 
nate> participant to profile a set of predictive algorithms and choose the 
nate> best.  Some good background info about this is here if someone wants to 
nate> take up this task:

nate> http://wikitest.freebsd.org/moin.cgi/powerd

nate> Another mitigating factor is a patch I hope to commit soon that removes 
nate> levels that aren't useful.  The general idea is the same as a recent 
nate> email from Tijl Coosemans but my approach is different.

I'm looking foaward to the effort in this area.

Sincerely,

--
Hajimu UMEMOTO @ Internet Mutual Aid Society Yokohama, Japan
ume at mahoroba.org  ume@{,jp.}FreeBSD.org
http://www.imasy.org/~ume/