Path: utzoo!mnetor!tmsoft!torsqnt!news-server.csri.toronto.edu!cs.utexas.edu!sun-barr!decwrl!pa.dec.com!decuac!hussar.dco.dec.com!mjr From: mjr@hussar.dco.dec.com (Marcus J. Ranum) Newsgroups: comp.unix.internals Subject: NFS kludges? Message-ID: <1991Feb26.191744.9090@decuac.dec.com> Date: 26 Feb 91 19:17:44 GMT References: <14367@ulysses.att.c <469@appserv.Eng.Sun.COM> Organization: Digital Equipment Corp., Washington Ultrix Resource Center Lines: 17 >>And to Unix users, NFS is not stateless. What is rpc.lockd used for? > >Whew! Where did this come from? NFS is stateless. The locking gunk is >and has been a problematic aspect for some time. Indeed, one is tempted to point to the "locking gunk" and how problematic it is - that should be an indicator of how hard getting remote device states to work properly would be. Devices aren't stateless. Locks are stateful. Devices are stateful. It's a whole different kettle of fish. NFS works terrifically well for what it is intended to do. It was not intended as a solution for the distributed database problem, and if it had been, people would either a) still be writing it or b) complaining a lot about how slow it is. mjr. Brought to you by Super Global Mega Corp .com