Nice report, nice tool :-) But not that much data from Windows 2000... (when things get tough - using disk or network, there are no data on Windows 2000 - why? Is it too good or too bad?)
/RogerL PS Nice to see that you have not ignored Benno Senoners latencytest DS On m�ndagen den 8 april 2002 22.49, John Regehr wrote: > Hi all, > > I'm audio-ignorant so I mostly lurk on this list, but I do usually > follow discussions about scheduling latency and stuff like that. > > I thought some members of this list would be interested in a paper I'm > working on that's going to appear in the FREENIX track at the USENIX > Technical Conference in June. It's about Hourglass, a user-level tool > for measuring scheduling behavior. There are some comparisons in the > paper between some of the different real-time enhanced versions of > Linux. > > There's a draft here: > http://www.cs.utah.edu/~regehr/hourglass-040802.ps > http://www.cs.utah.edu/~regehr/hourglass-040802.pdf > > The final copy is due in about a week; I thought I'd post a copy now in > case people find any mistakes, since this will give me the opportunity > to correct them. > > NOTE: This draft will disappear a few days after this message is posted. > Anyone reading this message in the archives after April 16 2002 should > instead look for the final version of the paper at the Hourglass site: > > http://www.cs.utah.edu/~regehr/hourglass > > Thanks, > > John Regehr > Postdoc, School of Computing, University of Utah -- Roger Larsson Skellefte� Sweden
