From owner-freebsd-bugs Sun Nov 10 19:30:04 1996 Return-Path: owner-bugs Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id TAA08408 for bugs-outgoing; Sun, 10 Nov 1996 19:30:04 -0800 (PST) Received: (from gnats@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id TAA08399; Sun, 10 Nov 1996 19:30:02 -0800 (PST) Date: Sun, 10 Nov 1996 19:30:02 -0800 (PST) Message-Id: <199611110330.TAA08399@freefall.freebsd.org> To: freebsd-bugs Cc: From: Adam David Subject: Re: bin/1891: mountd fails to export Reply-To: Adam David Sender: owner-bugs@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk The following reply was made to PR bin/1891; it has been noted by GNATS. From: Adam David To: J Wunsch Cc: FreeBSD-gnats-submit@freebsd.org Subject: Re: bin/1891: mountd fails to export Date: Mon, 11 Nov 1996 03:27:31 GMT > > >Fix: > > > > Skip the host instead of skipping immediately to the next line > > every time get_host() returns an error. > > > > or: > > > > Document that it is a requirement for all hosts in /etc/exports > > to be defined in /etc/hosts in order to prevent failed exports to > > reachable hosts. > > > > or: > > > > Document that it is advisable to separate host lists topologically > > into separate lines in /etc/exports > > I vote for 2 (and 3). (Except, no reference to /etc/hosts, the valid > hostname can also come from DNS.) > > -- > cheers, J"org > > joerg_wunsch@uriah.heep.sax.de -- http://www.sax.de/~joerg/ -- NIC: JW11-RIPE Why is it such a good idea to discard the whole line (or netgroup) if only one host (of many) cannot be resolved? What is gained by preventing other valid hosts from being exported to? The single hostname may be invalid because of a removal of a single host from the network, or a rename of a single host. Why should all other hosts be denied service because hosts information changes in this limited way while the change has not been reflected in /etc/exports ? -- Adam David