From owner-freebsd-hackers@freebsd.org Mon Oct 5 17:58:56 2020 Return-Path: Delivered-To: freebsd-hackers@mailman.nyi.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2610:1c1:1:606c::19:1]) by mailman.nyi.freebsd.org (Postfix) with ESMTP id 27CF642B354 for ; Mon, 5 Oct 2020 17:58:56 +0000 (UTC) (envelope-from steffen@sdaoden.eu) Received: from sdaoden.eu (sdaoden.eu [217.144.132.164]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (4096 bits) server-digest SHA256) (Client did not present a certificate) by mx1.freebsd.org (Postfix) with ESMTPS id 4C4pHy6qJpz4HL2; Mon, 5 Oct 2020 17:58:54 +0000 (UTC) (envelope-from steffen@sdaoden.eu) Received: by sdaoden.eu (Postfix, from userid 1000) id D4C1116057; Mon, 5 Oct 2020 19:58:53 +0200 (CEST) Date: Mon, 05 Oct 2020 19:58:53 +0200 From: Steffen Nurpmeso To: Eric McCorkle Cc: Alan Somers , FreeBSD Hackers Subject: Re: Mounting encrypted ZFS datasets/GELI for users? Message-ID: <20201005175853.4OgAF%steffen@sdaoden.eu> In-Reply-To: <00dbfac0-6c6f-355e-c21b-db2cae3a87e4@metricspace.net> References: <8d467e98-237f-c6a2-72de-94c0195ec964@metricspace.net> <630f9133-4f67-92bd-41f9-fb04d985c159@metricspace.net> <00dbfac0-6c6f-355e-c21b-db2cae3a87e4@metricspace.net> Mail-Followup-To: Eric McCorkle , Alan Somers , FreeBSD Hackers User-Agent: s-nail v14.9.19 OpenPGP: id=EE19E1C1F2F7054F8D3954D8308964B51883A0DD; url=https://ftp.sdaoden.eu/steffen.asc; preference=signencrypt BlahBlahBlah: Any stupid boy can crush a beetle. But all the professors in the world can make no bugs. MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable X-Rspamd-Queue-Id: 4C4pHy6qJpz4HL2 X-Spamd-Bar: / Authentication-Results: mx1.freebsd.org; dkim=none; dmarc=none; spf=pass (mx1.freebsd.org: domain of steffen@sdaoden.eu designates 217.144.132.164 as permitted sender) smtp.mailfrom=steffen@sdaoden.eu X-Spamd-Result: default: False [-0.97 / 15.00]; ARC_NA(0.00)[]; NEURAL_HAM_MEDIUM(-0.99)[-0.987]; FROM_HAS_DN(0.00)[]; RCPT_COUNT_THREE(0.00)[3]; R_SPF_ALLOW(-0.20)[+a:c]; NEURAL_HAM_LONG(-1.04)[-1.041]; MIME_GOOD(-0.10)[text/plain]; DMARC_NA(0.00)[sdaoden.eu]; TO_MATCH_ENVRCPT_SOME(0.00)[]; TO_DN_ALL(0.00)[]; NEURAL_HAM_SHORT(-0.64)[-0.639]; MID_CONTAINS_FROM(1.00)[]; RCVD_COUNT_ZERO(0.00)[0]; FROM_EQ_ENVFROM(0.00)[]; R_DKIM_NA(0.00)[]; MIME_TRACE(0.00)[0:+]; ASN(0.00)[asn:15987, ipnet:217.144.128.0/20, country:DE]; SUBJECT_ENDS_QUESTION(1.00)[]; MAILMAN_DEST(0.00)[freebsd-hackers] X-BeenThere: freebsd-hackers@freebsd.org X-Mailman-Version: 2.1.33 Precedence: list List-Id: Technical Discussions relating to FreeBSD List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 05 Oct 2020 17:58:56 -0000 Eric McCorkle wrote in <00dbfac0-6c6f-355e-c21b-db2cae3a87e4@metricspace.net>: |On 10/5/20 11:50 AM, Alan Somers wrote: |> On Mon, Oct 5, 2020 at 9:40 AM Eric McCorkle > wrote: |>=20 |> On 10/5/20 11:12 AM, Alan Somers wrote: |>=20 |>> First of all, what kind of thread are you concerned with?=C2=A0 Disk |>> encryption does not protect against an attacker with access to a live |>> machine; it only protects against an attacker with access to an off |>> machine, or to the bare HDDs.=C2=A0 Per-user encryption would presumab= ly |>> protect one user from another user who has physical access to the off |>> server.=C2=A0 Is that what you're worried about?=C2=A0 If not, then you |> shouldn't |>> bother with per-user encryption.=C2=A0 Just encrypt all of /home or al= l of |>> the pool with a single key. |>> |>> -Alan |>=20 |> I am evaluating options for domains where use of per-user encryptio= n \ |> is |> mandated, often as a means of protecting against insider threats. |>=20 |>=20 |> But if the victim user and the aggressor user are logged in at the same |> time, then both users' home directories will be decrypted, and unix |> permissions will be the only thing protecting the victim, right?=C2=A0 = That |> situation doesn't sound any better than no encryption at all.=C2=A0 And |> insiders who have offline access to the HDDs would be thwarted by global |> encryption just as much as per-user encryption.=C2=A0 I'm not denying t= hat |> you may be under some legal mandate for per-user encryption; I just |> don't understand the motivation. | |Per-user encryption is not perfect, but that's not the goal of |requirements like this. First of all, this can be used to protect |secure workstations, where it's reasonable to expect only one person to |be logged in at a time. | |Beyond that, the goal is to shrink the window of possible attacks and to |aid detection. If the Adversary has to be active while a particular |user is logged in, then they have a much smaller window of attack. |Moreover, this helps with forensics, as you can look at what else was |going on in the system in the much shorter window while a compromised |user was active. That project is very cool. I also want to thank for importing ZFS with encryption, i am not using it yet, but am looking forward to it. One important aspect of such (additional, on top of block encrypted disks) per-user-home encryption is that you can simply backup the entire directory without additional protection, if you have access to the unmounted content. I personally use several different encrypted directories, not the /home/steffen as such but sec.arena and sic therein, which get only mounted as necessary, and automatically unmounted (for all users) when the LID is closed. --steffen | |Der Kragenbaer, The moon bear, |der holt sich munter he cheerfully and one by one |einen nach dem anderen runter wa.ks himself off |(By Robert Gernhardt)