On Thu, Aug 19, 1999 at 10:39:01PM +0100, Paul Jakma wrote: > On Thu, 19 Aug 1999, Jan Edler wrote: > > It isn't reporting 99% of 4 Xeons, but 99% of one Xeon. > Bonnie just takes cpu/elapsed*100. > > it also depends on the load. if bonnie is the only real load on the > machine then cpu util will be 99%. Not in my experience. I see cpu load (calculated by bonnie) generally around N% when I see read througput of N MB/s. I'm sure the near equality is a coincidence, but the fact that the cpu load scales with the througput is significant. It says the cpu does work proportional to the amount of data being read, which is what it should be. Jan Edler NEC Research Institute
- 4 cheetah-3's pretty much saturate 80MB/sec channel James Manning
- Re: 4 cheetah-3's pretty much saturate 80MB/sec ch... Paul Jakma
- Re: 4 cheetah-3's pretty much saturate 80MB/se... James Manning
- Re: 4 cheetah-3's pretty much saturate 80MB/sec ch... Jan Edler
- Re: 4 cheetah-3's pretty much saturate 80MB/se... James Manning
- Re: 4 cheetah-3's pretty much saturate 80M... Jan Edler
- Re: 4 cheetah-3's pretty much saturate... Paul Jakma
- Re: 4 cheetah-3's pretty much sat... Jan Edler
- Re: 4 cheetah-3's pretty much sat... James Manning
- Re: 4 cheetah-3's pretty much saturate... James Manning
- Re: 4 cheetah-3's pretty much saturate 80M... Gadi Oxman
- Re: 4 cheetah-3's pretty much saturate 80MB/sec ch... Terje Marthinussen
- Re: 4 cheetah-3's pretty much saturate 80MB/sec ch... James Manning