Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 6 Feb 2006 14:14:21 -0700
From:      "Chad Leigh -- Shire.Net LLC" <chad@shire.net>
To:        =?ISO-8859-1?Q?Bj=F6rn_K=F6nig?= <bkoenig@cs.tu-berlin.de>
Cc:        current@freebsd.org
Subject:   Re: unprivileged users are able to kill certain jailed processes
Message-ID:  <778A6B9C-DADC-45AE-A5C8-DEFC2D2C41D4@shire.net>
In-Reply-To: <43E7B1A7.8010501@cs.tu-berlin.de>
References:  <43E60708.9000902@cs.tu-berlin.de> <43E7494B.9040401@freebsd.org> <43E7B1A7.8010501@cs.tu-berlin.de>

next in thread | previous in thread | raw e-mail | index | archive | help

On Feb 6, 2006, at 1:29 PM, Bj=F6rn K=F6nig wrote:

> Andre Oppermann schrieb:
>
>> [...] If you have normal users on the host and
>> have jails under the same user id then, yea, tough luck.  You're not
>> supposed to do that. [...]
>
> Yes, I can prevent from overlapping UIDs, but how to prevent from =20
> that if host administrator and jail administrator are two =20
> independent parties? It requires much more carefulness and =20
> precautions.

Well, the host admin, when detailing services and responsibilities to =20=

the jail admin (I have a similar situation), can tell the jail admin =20
which range of UIDs to use for new users.  I typically use the last =20
byte of the IP address * 100 as the base.

Eg, say a jail is 192.168.1.100 then they can start with 10000 as a =20
UID and go up to 10100.

Additionally, the host should ideally have no users but the bare =20
minimum for the admin.  All the "host"-based users and services =20
should ideally be in their own jail.

And if you can use a common base jail install mounted read only =20
inside each jail, you will greatly increase security of the jails as =20
exploits that replace system binaries will fail.

gruss aus utah
Chad


---
Chad Leigh -- Shire.Net LLC
Your Web App and Email hosting provider
chad at shire.net






Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?778A6B9C-DADC-45AE-A5C8-DEFC2D2C41D4>