Date: Wed, 7 Mar 2001 07:49:17 -0800 From: Robert Clark <res03db2@gte.net> To: Terry Lambert <tlambert@primenet.com> Cc: Wes Peters <wes@softweyr.com>, rjesup@wgate.com, Mike Meyer <mwm@mired.org>, Matt Dillon <dillon@earth.backplane.com>, Alfred Perlstein <bright@wintelcom.net>, josb@cncdsl.com, chat@FreeBSD.ORG Subject: Re: DJBDNS vs. BIND Message-ID: <20010307074917.B47638@darkstar.gte.net> In-Reply-To: <200103070840.BAA14141@usr05.primenet.com>; from tlambert@primenet.com on Wed, Mar 07, 2001 at 08:40:01AM %2B0000 References: <3AA5DB60.86A5C03D@softweyr.com> <200103070840.BAA14141@usr05.primenet.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, Mar 07, 2001 at 08:40:01AM +0000, Terry Lambert wrote: > > > I would argue that human recovery is not a useful scenario, even > > > in the absence of a backup. > > > > Which flies in the face of every system recovery ever attempted, including > > the one I got to do last week. Even if you just finished a full backup > > of the system when it crashed/got killed, some files may be out of date. > > You are thinking about systems which have sufficient exposed > complexity that there are likely to be operators on hand to > do the job; the incremental costs of doing the job are, I > think, unrelated. A storage format and appropriate tools to > allow severable partial recovery of the data by a human are > generally enough. > > Basically, this means that binary data is not the issue, easy > human recovery in this situation is. > > I'd also argue that this situtation itself is increasingly > rare. In an embedded system running FreeBSD, for example, the > only time an operator with the necessary capability will see > te data, one way or another, is in a post mortem of a returned > system. This means that the vast majority of cases require > the ability to perform automatic "best guess" recovery, at a > minimum, or "last change state rollback" (effectively, working > configuration versioning), at best. > > > Terry Lambert > terry@lambert.org > --- > Any opinions in this posting are my own and not those of my present > or previous employers. > In the instance of a post mortem, would a log of successive configuration changes be more valuable than a file where only the last state is stored? Some config files do track successive changes to a small degree. Or at least show something about what added a section to the file. (rc.conf) [RC] To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-chat" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20010307074917.B47638>