jpc@avdms8.msfc.nasa.gov (J. Porter Clark) (12/08/90)
The question I asked, stripped to the essentials, was: >Is there some reason that biod and nfsd never seem to get swapped out? Since the clue to the answer is in the manual, I apologize for asking! All of the responses I received were more or less along the lines of the first one I received, which was from tar@math.ksu.edu (Tim Ramsey): >nfsd and biod run as system processes. The /usr/etc/nfsd program >just opens a UDP socket, registers it with the portmapper, and invokes >the nfssvc(2) system call. That system call never returns (under normal >circumstances). The man page for nfssvc(2) under 4.1 says: > > Both system calls result in kernel-only processes with > user memory discarded. > >Since kernel memory is never swapped, nfsd and biod can't be swapped. To date, I have also received similar data from: ekrell@ulysses.att.com (Eduardo Krell) davide%cadillac.cad.mcc.com@mcc.com (David Eckelkamp) convex!datri@uxc.cso.uiuc.edu (Anthony A. Datri) Thanks to all who responded! Porter Clark NASA/MSFC Communications Systems Branch EB33 Huntsville, AL 35812 Phone: (205)544-3661 FAX: (205)544-9582 Internet: jpc@avdms8.msfc.nasa.gov