Return to BSD News archive
Path: euryale.cc.adfa.oz.au!newshost.anu.edu.au!harbinger.cc.monash.edu.au!news.mel.connect.com.au!munnari.OZ.AU!news.ecn.uoknor.edu!news.cis.okstate.edu!news.ksu.ksu.edu!news.physics.uiowa.edu!math.ohio-state.edu!uwm.edu!vixen.cso.uiuc.edu!newsfeed.internetmci.com!usenet.eel.ufl.edu!col.hp.com!sdd.hp.com!hamblin.math.byu.edu!park.uvsc.edu!usenet From: Terry Lambert <terry@lambert.org> Newsgroups: comp.unix.bsd.netbsd.misc,comp.unix.bsd.bsdi.misc,comp.unix.solaris,comp.unix.aix Subject: Re: ISP hardware/software choices (performance comparison) Date: 15 Jan 1996 23:58:20 GMT Organization: Utah Valley State College, Orem, Utah Lines: 63 Distribution: inet Message-ID: <4depms$bi5@park.uvsc.edu> References: <4cmopu$d35@vixen.cso.uiuc.edu> <4d43bt$es8@park.uvsc.edu> <4d6v8e$b1e@engnews2.Eng.Sun.COM> <4d9has$qo9@park.uvsc.edu> <4de3db$n6a@engnews2.Eng.Sun.COM> NNTP-Posting-Host: hecate.artisoft.com Xref: euryale.cc.adfa.oz.au comp.unix.bsd.netbsd.misc:1897 comp.unix.bsd.bsdi.misc:2042 comp.unix.solaris:56929 comp.unix.aix:68470 thurlow@peyto.eng.sun.com (Robert Thurlow) wrote: ] ] In article <4d9has$qo9@park.uvsc.edu>, ] Terry Lambert <terry@lambert.org> wrote: ] ] >thurlow@peyto.eng.sun.com (Robert Thurlow) wrote: ] ] >] NFS in 5.x has the same or better consistency guarantees as your ] >] favorite 4.x version. And in 5.5, you get NFS Version 3 and NFS ] >] over TCP, which work so well and so fast I was amazed when I ] >] started here. You're clearly out in left field on this topic. ] ] >Sorry, "the same" is not an argument for change. If you had a ] >tangible "or better", then you'd have a case. ] ] You implied that SunOS 5.x was faster because it played fast and ] loose with consistency guarantees. This is simply untrue, and ] I'm calling you on it; consistency has been tightened up in many ] areas, and not made weaker anywhere. If you believe otherwise, ] please be specific. Client caching prior to NFSv3 violates the protocol specification. Server caching of writes violates the protocol specification. 4.x did not do server write caching by default. I will not argue that NFS security under 5.x isn't tighter (and therefore less usable) than under 4.x. Most of the "consitency guarantees" I see in the 5.x release notes are security issues, and not relevant to the discussion. The one possible win (and you have yet to claim it, or anything other than a blanket statement that 5.x is faster than 4.x, without providing numbers or rationale) is kernel threading of the biod's. This is a minor win at best, since the process context switch on NFS is not that high, relatively, because the anonymity of clients breaks FS locality of refernce, which renders the cache and register windows useless. Thus effective context switch overhead vs. a system-call-that-never-returns is negligible. Must I both make and refute your points? It is encumbent on *you* to defend your speed claims with specific examples. To get this off on the right foot: ] >] NFS in 5.x has the same or better consistency guarantees as your ] >] favorite 4.x version. And in 5.5, you get NFS Version 3 and NFS ] >] over TCP, which work so well and so fast I was amazed when I ] >] started here. You're clearly out in left field on this topic. I don't believe you. Please give specific references. Regards, Terry Lambert terry@cs.weber.edu --- Any opinions in this posting are my own and not those of my present or previous employers.