Firejail is a SUID sandbox program that reduces the risk of security breaches by restricting the running environment of untrusted applications using Linux namespaces. It allows a process and all its descendants to have their own private view of the globally shared kernel resources, such as the network stack, process table, mount table.
Started as a simple sandbox for Mozilla Firefox, Firejail was expanded to work on any type of executable, such as servers, graphic programs, and even as login shell.
The program is written in C and only needs libc and POSIX threads (libpthreads), available by default on any Linux platform. The download page provides source code (./configure && make && sudo make install), deb (dpkg -i firejail.deb) and rpm (rpm -i firejail.rpm) packages. Once installed, you can start a program in sandbox as:
$ firejail [options] program and arguments
$ firejail --debug firefox
To login into a Firejail sandbox, you need to set /usr/bin/firejail as user shell in /etc/passwd. You can change the shell for an existing user with chsh command (chsh –shell /usr/bin/firejail), or you can use adduser command to define the shell when the user account is created (adduser –shell /usr/bin/firejail username). Below is a ssh login session into a sandboxed account:
SSH login into a default Firejail sandbox
The traditional Linux security model starts with file permissions. The model lets the kernel decide whether or not a process may access a resource based on permissions set as part of the filesystem. The coarse-grained granularity of this model often causes Linux processes to have too many rights. If more granularity is needed, one has to resort to adding security related code into the program source.
This series of articles is about Linux namespaces, a lightweight virtualization technology implemented in Linux kernel. In part 1 I’ve talked about building chroot jails using mount namespace, and in part 2 I’ve looked into isolating processes using PID namespace. The next step is to isolate the TCP/IP networking stack using network namespaces.
Security at this level is always reactive. Assuming the bad guy breaks into your server, he will realize he doesn’t have root privileges (classic Unix privilege separation implemented in server software), he runs on top of a fake filesystem (chroot), and he cannot get outside on the network. The later is usually done by placing the computer in a Demilitarized Zone (DMZ) behind a firewall.
The same effect can be achieved on the cheap using Linux namespaces. For this, I place the server in a container (vm1) running its own network segment (10.10.20.0/24). The container is connected to the host through a Linux bridge interface (br0). On the host I configure iptables firewall, isolating the server and effectively limiting the potential damage that could be inflicted on the larger network. The final setup looks like this:
Linux ptrace() system call provides a means by which one process may observe and control the execution of another process. It is primarily used to implement breakpoint debugging with gdb and system call tracing with strace. In this article I will look at the security implications of ptrace, and how to overcome them using Linux PID namespaces.
Public enemy numero uno
This is an experiment every Linux enthusiast should try. Start an ssh connection in a terminal and stop when you are just about to enter the password:
Starting an ssh session
Open another terminal, find the pid of your ssh session (ps ax | grep ssh) and start strace on it (strace -p 3660). Then, go back to your ssh terminal, type in your password, and watch it flying across strace terminal:
ptrace usage example
Namespace isolation is the simplest virtualization technology available in Linux kernel. It allows a process and all its descendants to have their own private view of the globally shared kernel resources, such as the network stack, process table, mount table. This feature is mostly popularized and promoted by utilities such as LXC (Linux Containers), Docker and virtenv.
Three syscalls are used to create Linux namespaces, unshare(), clone() and setns(). In this article I will take a look at unshare() and show how to use it directly in your scripts and programs without going through LXC or any other higher level virtualization tool.
I’ll start by investigating unshare command available in util-linux package, and from there I’ll move to the system call. In the end I’ll build a small C program that isolates a web browser such as Mozilla Firefox into a kernel namespace.
Virtualization refers to the creation of virtual machines that acts like real computers with an operating system. Software executed on these virtual machines is separated from the underlying hardware resources.
This article discusses LXC, a lightweight virtualization technology built into Linux kernel. The user space LXC tool is distributed with a number of templates that allow the creation of different Linux distro filesystems, usually one template for each major Linux distribution. The problem with these templates is they never work, or they stop working with every new release of LXC tool or of the particular Linux distribution. This is the case with all Linux distributions, and Debian is no exception. Currently, the Debian template is borken under “wheezy”. The relevant Debian bug is here, and history shows that as soon such a bug gets fixed, lxc user space driver changes again and breaks it. It could be worse, in Fedora LXC was broken in Fedora 15 and it was never fixed.
The simple way to handle the problem is to forget all about the template mechanism and roll your own containers. In Debian you can build the container filesystem using the standard debootstrap, or mount read-only the host filesystem, and then use lxc-execute to start a simple bash session inside the container. In this session you can than start all the programs you need to run in the container. It is an application container, very similar to the containers created using the official ssh template distributed with LXC.
The virtual machine I will describe in this article uses a root filesystem build using debootstrap (apt-get install debootstrap). The procedure is simple and it should work on any Debian machine. It will probably work also on any other distro based on Debian, such as Ubuntu, Mint etc.
Virtualization allows the creation of multiple virtual machines (VM) on top of an existing computer, each VM configured in a very specific way. All virtual machines run in parallel alongside the regular host applications, without affecting the host system. The type of virtualization I am currently using is Linux containers (LXC), a lightweight virtualization technology built into Linux kernel.
This is my third Debian virtualization article. In the first article, I’ve described the steps to create and run a basic virtual machine using LXC. In the second article I’ve isolated the VM on its own network segment, with its own TCP/IP networking stack. Both articles were dealing with server VMs, I’ve used Lighttpd as an example throughout my articles.
I will describe now how to run desktop applications such as Mozilla Firefox and LibreOffice in a virtual machine. I will use virtenv to build and run the VM. virtenv is a QT4 application released under GPLv2 license. It is basically a configuration wizard that allows you to configure and start the LXC-based virtual machines. Once the VM is started, you get a regular controlling terminal (xterm) and a desktop window running the lightweight Openbox window manager. In this window you can run your GUI applications, very similar to VMware Workstation or Oracle’s VirtualBox.
virtenv desktop and controlling terminal
Linux containers (LXC) is a lightweight virtualization technology built into Linux kernel. In my previous article, Debian Virtualization: LXC Application Containers, I have detailed the steps to configure and run a simple application container using LXC. LXC application containers are very lean and consume strictly the resources the application requires. This is in sharp contrast with other virtualization technologies which are running a full Linux distribution in VM.
The container uses its own file system, built by mounting read-only the relevant directories from the host file system. The host is an older computer running Debian 7 “wheezy”. The virtual machine is controlled through GNU screen if the VM was started automatically at boot time, or through a regular xterm.
One thing I left out was the networking stack. In my Lighttpd web server example, the VM uses the same networking stack as the host. This could become a problem if someone manages to compromise the web server: the intruder could then probe the networks connected to our host, in search for the next victim.
In this article I’ll modify the VM to run on a separate networking stack. I will place the VM on its own network segment, connected to the host through a Linux bridge interface. I will then go and set up the host firewall using iptables. This effectively isolates the VM and limits the potential damage that could be inflicted on the larger network. The final setup looks like this:
Linux containers (LXC) is a lightweight virtualization technology built into Linux kernel. Unlike other similar technologies, the virtual machines (VM) are driven without any overhead by the kernel already running on the computer. In VM you run only the processes you need, most of the time without even going through the regular SysV or Linux init. This means that memory is used very conservatively. These lightweight containers are sometimes called application containers, as opposed to distribution containers where you run a full distro starting with SysV/Linux init.
In this article I’ll take a look at installing and building a small web server application container using LXC on a Debian 7 workstation. Debian is a popular distribution for personal use among software developers, and the amount of software packaged far exceeds any other Linux disto out there. It also encourages tinkering, experimenting, and in a more general sense, learning about Linux and Free/Open Source Software.
The computer I am using is an old amd64 dual-core computer with a minimal LXDE desktop manager installed. For those of you interested, I have detailed the installation steps in my Lightweight Debian: LXDE Desktop From Scratch article. Since we are dealing with kernel utilities, all the commands in this article are specified as user root.
For my memory comparison of light Linux desktops I needed a tool that would allow me to install on my computer about 20 window managers/desktop environments. After looking at several common virtualization packages, I ended up using Linux containers and virtenv for the job.
LXC and virtenv
Probably the best way to describe virtenv is as a graphic interface for Linux containers utilities developed and distributed by LXC project. Linux containers is the virtualization technology build into Linux kernel, available in any kernel after 2.6.32.
The virtual machines (VM) are driven without any overhead by the kernel already running on the computer. You don’t need to run a different kernel in the virtual machine, run only the processes you need, without even going trough the regular SysV or Linux init. This all means that memory is used very conservatively. For example, on a 1GB RAM computer you can run easily 10 SSH/DHCP servers, or 10 different xorg/X11 servers with LXDE window managers on top.
ezchroot is a small script to chroot into OpenVZ containers. Once inside, you can update or modify the container software. The operation is similar to ezlxc.
if [ $# -gt 0 ]; then
echo "Usage: ezchroot directory"
cp -L /etc/resolv.conf $1/etc/.
mount -t proc none $1/proc
mount --rbind /dev $1/dev
mount --rbind /sys $1/sys
echo "entering chroot directory"
env NAME=chroot chroot $1 /bin/bash
echo "chroot exited"