It really depends on the definition of idl. Is idl defined to be the rounded idle time or 100 - (sum of the rounded sys + rounded usr). They will be different (as in the example below). Which is correct is dependent on the definition you want to use.
But as Jim points out, for most analysis the difference is not likely to be significant. As Emerson said "A foolish consistency is the hobgoblin of little minds." Dave Jim Mauro wrote, On 01/28/09 10:19: > So that was you! I never know it was you that filed the wt bug > (but I should have guessed! :^). > > I actually agree that the "adding up to 100" really does not matter. > In terms of the practical use of CPU utilization data, I don't really > distinguish between 90% idle and 88% idle, or %sys at 50% versus > 54%, etc. Plus or minus 2% for any given CPU state doesn't change > anything in terms of the usefullness of the metric and how I use it. > > Thanks, > /jim > > > Phil Harman wrote: >> Old news, I'm afraid. Any consumer of the cpu sys kstats will seem the >> same thing. I guess some people just didn't notice until we started >> discussing it :) >> >> At least I managed to get wt permanently set to zero ... >> >> 4518644 I/O wait statistic is still misleading and should be dropped >> >> http://bugs.opensolaris.org/view_bug.do?bug_id=4518644 >> >> Actually, I did try to address this adding up to 100 issue at about the >> same time, but I was told it didn't matter. Of course now that we have >> OpenSolaris there's nothing to stop someone who cares enough to propose >> a fix. >> >> Phil >> >> >> >> Stefan Parvu wrote: >> >>> CPU minf mjf xcal intr ithr csw icsw migr smtx srw syscl usr sys wt idl >>> 0 84 0 0 392 192 706 101 0 0 1 3414 10 3 0 86 >>> 0 5 0 0 308 108 230 9 0 0 0 422 1 1 0 97 >>> 0 0 0 0 308 108 232 9 0 0 0 416 1 1 0 97 >>> 0 1 0 0 311 111 239 10 0 0 0 470 2 2 0 97 >>> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ >>> 0 0 0 0 307 107 234 9 0 0 0 413 1 2 0 97 >>> 0 0 0 0 347 147 444 14 0 0 0 1104 2 2 0 96 >>> 0 0 0 0 346 146 290 9 0 0 0 451 1 2 0 97 >>> >>> It seems this happens as well on mpstat. Hmmm, could be something wider >>> broken ? >>> >>> My system: >>> Solaris Express Community Edition snv_103 X86 >>> Copyright 2008 Sun Microsystems, Inc. All Rights Reserved. >>> Use is subject to license terms. >>> Assembled 17 November 2008 >>> >>> System Configuration: Acer, inc. Ferrari 4000 >>> BIOS Configuration: Acer 3A27 03/20/06 >>> >>> Thanks, >>> stefan >>> >>> _______________________________________________ >>> perf-discuss mailing list >>> perf-discuss@opensolaris.org >>> >>> >> _______________________________________________ >> perf-discuss mailing list >> perf-discuss@opensolaris.org >> > _______________________________________________ > perf-discuss mailing list > perf-discuss@opensolaris.org _______________________________________________ perf-discuss mailing list perf-discuss@opensolaris.org
