The primary reason is to have a firewall that will be always up-to-update, unlike most commercial customer grade firewalls that are only supported for a few years. Having a firewall that runs opensource software - it’s based on FreeBSD - also make it easier to review and to verify that there are no back doors.
When I tried to upgrade it to the latest release - 19.1.7 - the upgrade failed because the filesystem ran out of inodes. There is already a topic about this at the OPNsense forum and a fix available for the upcoming nano OPNsense images.
But this will only resolve the issue when a new image becomes available and would require a reinstallation of the firewall.
Create a backup with the old-school dd, just in case.
123456
root@freebsd:~ # dd if=/dev/da0 of=/home/staf/opnsense.dd bs=4M status=progress
4013948928 bytes (4014 MB, 3828 MiB) transferred 415.226s, 9667 kB/s
957+1 records in
957+1 records out
4017807360 bytes transferred in 415.634273 secs (9666689 bytes/sec)
root@freebsd:~ #
mount
Verify that we can mount the filesystem.
1234567891011121314151617
root@freebsd:/ # mount /dev/da0a /mnt
root@freebsd:/ # cd /mnt
root@freebsd:/mnt # ls
.cshrc dev net
.probe.for.install.media entropy proc
.profile etc rescue
.rnd home root
COPYRIGHT lib sbin
bin libexec sys
boot lost+found tmp
boot.config media usr
conf mnt var
root@freebsd:/mnt # cd ..
root@freebsd:/ # df -i /mnt
Filesystem 1K-blocks Used Avail Capacity iused ifree %iused Mounted on
/dev/da0a 3916903 1237690 2365861 34% 38203 6595 85% /mnt
root@freebsd:/ #
umount
1
root@freebsd:/ # umount /mnt
dump
We’ll a create a backup with dump, we’ll use this backup to restore it again after we created a new filesystem with enough inodes.
123456789101112131415161718192021
root@freebsd:/ # dump 0uaf /home/staf/opensense.dump /dev/da0a
DUMP: Date of this level 0 dump: Sun May 19 10:43:56 2019
DUMP: Date of last level 0 dump: the epoch
DUMP: Dumping /dev/da0a to /home/staf/opensense.dump
DUMP: mapping (Pass I) [regular files]
DUMP: mapping (Pass II) [directories]
DUMP: estimated 1264181 tape blocks.
DUMP: dumping (Pass III) [directories]
DUMP: dumping (Pass IV) [regular files]
DUMP: 16.21% done, finished in 0:25 at Sun May 19 11:14:51 2019
DUMP: 33.27% done, finished in 0:20 at Sun May 19 11:14:03 2019
DUMP: 49.65% done, finished in 0:15 at Sun May 19 11:14:12 2019
DUMP: 66.75% done, finished in 0:09 at Sun May 19 11:13:57 2019
DUMP: 84.20% done, finished in 0:04 at Sun May 19 11:13:41 2019
DUMP: 99.99% done, finished soon
DUMP: DUMP: 1267205 tape blocks on 1 volume
DUMP: finished in 1800 seconds, throughput 704 KBytes/sec
DUMP: level 0 dump on Sun May 19 10:43:56 2019
DUMP: Closing /home/staf/opensense.dump
DUMP: DUMP IS DONE
root@freebsd:/ #
newfs
According to the newfs manpage:
We can specify the inode density with the -i option. A lower number will give use more inodes.
1234567
root@freebsd:/ # newfs -i 1 /dev/da0a
density increased from 1 to 4096
/dev/da0a: 3831.7MB (7847280 sectors) block size 32768, fragment size 4096
using 8 cylinder groups of 479.00MB, 15328 blks, 122624 inodes.
super-block backups (for fsck_ffs -b #) at:
192, 981184, 1962176, 2943168, 3924160, 4905152, 5886144, 6867136
root@freebsd:/ #
The installation had a OPNsense_Nano label, underscores are not allowed anymore in label name on FreeBSD 12.
So I used OPNsense instead, we’ll update /etc/fstab with the new label name.
In my previous two posts (1,
2 ), we created Docker Debian and
Arch-based images from scratch for the i386 architecture.
In this blog post - last one in this series - we’ll do the same for yum based distributions like CentOS and Fedora.
Building your own Docker base images isn’t difficult and let you trust your distribution Gpg signing keys instead of the docker hub. As explained in the first blog post. The mkimage scripts in the contrib directory of the Moby project git repository is a good place to start if you want to build own docker images.
[staf@centos386 github]$ cd moby/contrib/
[staf@centos386 contrib]$
mkimage-yum.sh
When you run mkimage-yum.sh you get the usage message.
123456789101112
[staf@centos386 contrib]$ ./mkimage-yum.sh
mkimage-yum.sh [OPTIONS] <name>
OPTIONS:
-p "<packages>" The list of packages to install in the container.
The default is blank. Can use multiple times.
-g "<groups>" The groups of packages to install in the container.
The default is "Core". Can use multiple times.
-y <yumconf> The path to the yum config to install packages from. The
default is /etc/yum.conf for Centos/RHEL and /etc/dnf/dnf.conf for Fedora
-t <tag> Specify Tag information.
default is reffered at /etc/{redhat,system}-release
[staf@centos386 contrib]$
build the image
The mkimage-yum.sh script will use /etc/yum.conf or /etc/dnf.conf to build the image. mkimage-yum.sh <name> will create the image with name.
1234567891011121314151617181920212223242526272829
[staf@centos386 contrib]$ sudo ./mkimage-yum.sh centos
[sudo] password for staf:
+ mkdir -m 755 /tmp/mkimage-yum.sh.LeZQNh/dev
+ mknod -m 600 /tmp/mkimage-yum.sh.LeZQNh/dev/console c 5 1
+ mknod -m 600 /tmp/mkimage-yum.sh.LeZQNh/dev/initctl p
+ mknod -m 666 /tmp/mkimage-yum.sh.LeZQNh/dev/full c 1 7
+ mknod -m 666 /tmp/mkimage-yum.sh.LeZQNh/dev/null c 1 3
+ mknod -m 666 /tmp/mkimage-yum.sh.LeZQNh/dev/ptmx c 5 2
+ mknod -m 666 /tmp/mkimage-yum.sh.LeZQNh/dev/random c 1 8
+ mknod -m 666 /tmp/mkimage-yum.sh.LeZQNh/dev/tty c 5 0
+ mknod -m 666 /tmp/mkimage-yum.sh.LeZQNh/dev/tty0 c 4 0
+ mknod -m 666 /tmp/mkimage-yum.sh.LeZQNh/dev/urandom c 1 9
+ mknod -m 666 /tmp/mkimage-yum.sh.LeZQNh/dev/zero c 1 5
+ '[' -d /etc/yum/vars ']'
+ mkdir -p -m 755 /tmp/mkimage-yum.sh.LeZQNh/etc/yum
+ cp -a /etc/yum/vars /tmp/mkimage-yum.sh.LeZQNh/etc/yum/
+ [[ -n Core ]]
+ yum -c /etc/yum.conf --installroot=/tmp/mkimage-yum.sh.LeZQNh --releasever=/ --setopt=tsflags=nodocs --setopt=group_package_types=mandatory -y groupinstall Core
Loaded plugins: fastestmirror, langpacks
There is no installed groups file.
Maybe run: yum groups mark convert (see man yum)
<snip>
+ tar --numeric-owner -c -C /tmp/mkimage-yum.sh.LeZQNh .
+ docker import - centos:7.6.1810
sha256:7cdb02046bff4c5065de670604fb3252b1221c4853cb4a905ca04488f44f52a8
+ docker run -i -t --rm centos:7.6.1810 /bin/bash -c 'echo success'
success
+ rm -rf /tmp/mkimage-yum.sh.LeZQNh
[staf@centos386 contrib]$
Rename
A new image is created with the name centos.
1234
[staf@centos386 contrib]$ docker images
REPOSITORY TAG IMAGE ID CREATED SIZE
centos 7.6.1810 7cdb02046bff 3 minutes ago 281 MB
[staf@centos386 contrib]$
You might want to rename to include your name or project name. You can do this by retag the image and remove the old image name.
12345678910111213141516171819
[staf@centos386 contrib]$ docker images
REPOSITORY TAG IMAGE ID CREATED SIZE
centos 7.6.1810 7cdb02046bff 20 seconds ago 281 MB
[staf@centos386 contrib]$ docker rmi centos
Error response from daemon: No such image: centos:latest
[staf@centos386 contrib]$ docker images
REPOSITORY TAG IMAGE ID CREATED SIZE
centos 7.6.1810 7cdb02046bff 3 minutes ago 281 MB
[staf@centos386 contrib]$ docker tag 7cdb02046bff stafwag/centos_386:7.6.1810
[staf@centos386 contrib]$ docker images
REPOSITORY TAG IMAGE ID CREATED SIZE
centos 7.6.1810 7cdb02046bff 7 minutes ago 281 MB
stafwag/centos_386 7.6.1810 7cdb02046bff 7 minutes ago 281 MB
[staf@centos386 contrib]$ docker rmi centos:7.6.1810
Untagged: centos:7.6.1810
[staf@centos386 contrib]$ docker images
REPOSITORY TAG IMAGE ID CREATED SIZE
stafwag/centos_386 7.6.1810 7cdb02046bff 8 minutes ago 281 MB
[staf@centos386 contrib]$
[staf@archlinux32 archlinux32]$ TMPDIR=`pwd` sudo ./mkimage-arch.sh
spawn pacstrap -C ./mkimage-arch-pacman.conf -c -d -G -i /var/tmp/rootfs-archlinux-wqxW0uxy8X base bash haveged pacman pacman-mirrorlist --ignore dhcpcd,diffutils,file,inetutils,iproute2,iputils,jfsutils,licenses,linux,linux-firmware,lvm2,man-db,man-pages,mdadm,nano,netctl,openresolv,pciutils,pcmciautils,psmisc,reiserfsprogs,s-nail,sysfsutils,systemd-sysvcompat,usbutils,vi,which,xfsprogs
==> Creating install root at /var/tmp/rootfs-archlinux-wqxW0uxy8X
==> Installing packages to /var/tmp/rootfs-archlinux-wqxW0uxy8X
:: Synchronizing package databases...
core 198.0 KiB 676K/s 00:00 [##########################################] 100%
extra 2.4 MiB 1525K/s 00:02 [##########################################] 100%
community 6.3 MiB 396K/s 00:16 [##########################################] 100%
:: dhcpcd is in IgnorePkg/IgnoreGroup. Install anyway? [Y/n] n
:: diffutils is in IgnorePkg/IgnoreGroup. Install anyway? [Y/n] n
:: file is in IgnorePkg/IgnoreGroup. Install anyway? [Y/n] n
<snip>
==> WARNING: /var/tmp/rootfs-archlinux-wqxW0uxy8X is not a mountpoint. This may have undesirable side effects.
Generating locales...
en_US.UTF-8... done
Generation complete.
tar: ./etc/pacman.d/gnupg/S.gpg-agent.ssh: socket ignored
tar: ./etc/pacman.d/gnupg/S.gpg-agent.extra: socket ignored
tar: ./etc/pacman.d/gnupg/S.gpg-agent: socket ignored
tar: ./etc/pacman.d/gnupg/S.gpg-agent.browser: socket ignored
sha256:41cd9d9163a17e702384168733a9ca1ade0c6497d4e49a2c641b3eb34251bde1
Success.
[staf@archlinux32 archlinux32]$
Rename
A new image is created with the name archlinux.
1234
[staf@archlinux32 archlinux32]$ docker images
REPOSITORY TAG IMAGE ID CREATED SIZE
archlinux latest 18e74c4d823c About a minute ago 472MB
[staf@archlinux32 archlinux32]$
You might want to rename it. You can do this by retag the image and remove the old image name.
12345678910111213141516
[staf@archlinux32 archlinux32]$ docker images
REPOSITORY TAG IMAGE ID CREATED SIZE
archlinux latest 18e74c4d823c About a minute ago 472MB
[staf@archlinux32 archlinux32]$ docker tag stafwag/archlinux:386 18e74c4d823c
Error response from daemon: No such image: stafwag/archlinux:386
[staf@archlinux32 archlinux32]$ docker tag 18e74c4d823c stafwag/archlinux:386
[staf@archlinux32 archlinux32]$ docker images
REPOSITORY TAG IMAGE ID CREATED SIZE
archlinux latest 18e74c4d823c 3 minutes ago 472MB
stafwag/archlinux 386 18e74c4d823c 3 minutes ago 472MB
[staf@archlinux32 archlinux32]$ docker rmi archlinux
Untagged: archlinux:latest
[staf@archlinux32 archlinux32]$ docker images
REPOSITORY TAG IMAGE ID CREATED SIZE
stafwag/archlinux 386 18e74c4d823c 3 minutes ago 472MB
[staf@archlinux32 archlinux32]$
Test
123456789
[staf@archlinux32 archlinux32]$ docker run --rm -it stafwag/archlinux:386 /bin/sh
sh-5.0# pacman -Syu
:: Synchronizing package databases...
core is up to date
extra is up to date
community is up to date
:: Starting full system upgrade...
there is nothing to do
sh-5.0#
I was using docker on an Odroid U3, but my Odroid stopped working. I switched to another system that is i386 only.
You’ll find my journey to build docker images for i386 below.
Reasons to build your own docker images
If you want to use docker you can start with docker images on the docker registry.
There are several reasons to build your own base images.
Security
The first reason is security, docker images are not signed by default.
Anyone can upload docker images to the public docker hub with bugs or malicious code.
There are “official” docker images available at https://docs.docker.com/docker-hub/official_images/ when you execute a docker search the official docker images are tagged on the official column and are also signed by Docker. To only allow signed docker images you need to set the DOCKER_CONTENT_TRUST=1 environment variable. - This should be the default IMHO -
There is one distinction, the “official” docker images are signed by the “Repo admin” of the Docker hub, not by the official GNU/Linux distribution project.
If you want to trust the official project instead of the Docker repo admin you can resolve this building your own images.
Support other architectures
Docker images are generally built for AMD64 architecture. If you want to use other architectures - ARM, Power, SPARC or even i386 - you’ll find some images on the Docker hub but these are usually not Official docker images.
Control
When you build your own images, you have more control over what goes or not goes into the image.
Building your own docker base images
There are several ways to build your own docker images.
The Mobyproject is Docker’s development project - a bit like what Fedora is to RedHat -.
The Moby project has a few scripts that help you to create docker base images and is also a good start if you want to review how to build your own images.
GNU/Linux distributions
I build the images on the same GNU/Linux distribution (e.g. The debian images are build on a Debian system) to get the correct gpg keys.
Debian GNU/Linux & Co
Debian GNU/Linux makes it very easy to build your own Docker base images. Only debootstrap is required.
I’ll use the moby script to the Debian base image and debootstrap to build an i386 docker Ubuntu 18.04 image.
Ubuntu doesn’t support i386 officially but includes the i386 userland so it’s possible to build i386 Docker images.
staf@whale:~/github/moby/contrib$ sudo apt install debootstrap
[sudo] password for staf:
Reading package lists... Done
Building dependency tree
Reading state information... Done
debootstrap is already the newest version (1.0.114).
debootstrap set to manually installed.
0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded.
staf@whale:~/github/moby/contrib$
The Moby way
Go to the contrib directory
12
staf@whale:~/github$ cd moby/contrib/
staf@whale:~/github/moby/contrib$
mkimage.sh
mkimage.sh --help gives you more details howto use the script.
staf@whale:~/github/moby/contrib$ docker images
REPOSITORY TAG IMAGE ID CREATED SIZE
stafwag/debian_i386 stretch cb96d1663079 About a minute ago 97.6MB
staf@whale:~/github/moby/contrib$
staf@ubuntu184:~/github/moby$ sudo apt install debootstrap
Reading package lists... Done
Building dependency tree
Reading state information... Done
Suggested packages:
ubuntu-archive-keyring
The following NEW packages will be installed:
debootstrap
0 upgraded, 1 newly installed, 0 to remove and 0 not upgraded.
Need to get 35,7 kB of archives.
After this operation, 270 kB of additional disk space will be used.
Get:1 http://be.archive.ubuntu.com/ubuntu bionic-updates/main amd64 debootstrap all 1.0.95ubuntu0.3 [35,7 kB]
Fetched 35,7 kB in 0s (85,9 kB/s)
Selecting previously unselected package debootstrap.
(Reading database ... 163561 files and directories currently installed.)
Preparing to unpack .../debootstrap_1.0.95ubuntu0.3_all.deb ...
Unpacking debootstrap (1.0.95ubuntu0.3) ...
Processing triggers for man-db (2.8.3-2ubuntu0.1) ...
Setting up debootstrap (1.0.95ubuntu0.3) ...
staf@ubuntu184:~/github/moby$
bootsrap
Create a directory that will hold the chrooted operating system.
You can customize your installation before it goes into the image. One thing that you should customize is include update in the image.
Update /etc/resolve.conf
1
staf@ubuntu184:~/dockerbuild/ubuntu$ sudo vi chroot-bionic/etc/resolv.conf
1
nameserver 9.9.9.9
Update /etc/apt/sources.list
1
staf@ubuntu184:~/dockerbuild/ubuntu$ sudo vi chroot-bionic/etc/apt/sources.list
And include the updates
123
deb http://ftp.ubuntu.com/ubuntu bionic main
deb http://security.ubuntu.com/ubuntu bionic-security main
deb http://ftp.ubuntu.com/ubuntu/ bionic-updates main
Chroot into your installation and run apt-get update
staf@ubuntu184:~/dockerbuild/ubuntu/chroot-bionic$ docker images
REPOSITORY TAG IMAGE ID CREATED SIZE
stafwag/ubuntu_i386 bionic 83560ef3c8d4 About a minute ago 315MB
staf@ubuntu184:~/dockerbuild/ubuntu/chroot-bionic$
12345678
staf@ubuntu184:~/dockerbuild/ubuntu/chroot-bionic$ docker run -it --rm stafwag/ubuntu_i386:bionic /bin/bash
root@665cec6ee24f:/# lsb_release -a
No LSB modules are available.
Distributor ID: Ubuntu
Description: Ubuntu 18.04.2 LTS
Release: 18.04
Codename: bionic
root@665cec6ee24f:/#
In a traditional environment, systems are installed from a CDROM. The configuration is executed by the system administrator through the installer. This soon becomes a borning and unpractical task when we need to set up a lot of systems also it is important that systems are configured in same - and hopefully correct - way.
In a traditional environment, this can be automated by booting via BOOTP/PXE boot and configured is by a system that “feeds” the installer. Examples are:
In a cloud environment, we use images to install systems. The system automation is generally done by cloud-init. Cloud-init was originally developed for Ubuntu GNU/Linux on the Amazon EC2 cloud. It has become the de facto installation configuration tool for most Unix like systems on most cloud environments.
Cloud-init uses a YAML file to configure the system.
Images
Most GNU/Linux distributions provide images that can be used to provision a new system.
You can find the complete list on the OpenStack website
[staf@centos7 iso]$ gpg --verify sha256sum.txt.asc
gpg: Signature made Thu 31 Jan 2019 04:28:30 PM CET using RSA key ID F4A80EB5
gpg: Good signature from "CentOS-7 Key (CentOS 7 Official Signing Key) <security@centos.org>"
gpg: WARNING: This key is not certified with a trusted signature!
gpg: There is no indication that the signature belongs to the owner.
Primary key fingerprint: 6341 AB27 53D7 8A78 A7C2 7BB1 24C6 A8A7 F4A8 0EB5
[staf@centos7 iso]$
The key fingerprint must match the one of RPM-GPG-KEY-CentOS-7.
Verify the iso file
1234
[staf@centos7 iso]$ xz -d CentOS-7-x86_64-GenericCloud-1901.qcow2.xz
[staf@centos7 iso]$ sha256sum -c sha256sum.txt.asc 2>&1 | grep OK
CentOS-7-x86_64-GenericCloud-1901.qcow2: OK
[staf@centos7 iso]$
Image
info
The image we download is a normal qcow2 image, we can see the image information with qemu-info
123456789
[root@centos7 iso]# qemu-img info CentOS-7-x86_64-GenericCloud-1901.qcow2
image: CentOS-7-x86_64-GenericCloud-1901.qcow2
file format: qcow2
virtual size: 8.0G (8589934592 bytes)
disk size: 895M
cluster_size: 65536
Format specific information:
compat: 0.10
[root@centos7 iso]#
Copy & resize
The default image is small - 8GB - we might be using the image to provision other systems so it better to leave it untouched.
Copy the image to the location where we’ll run the virtual system.
We’ll create a simple cloud-init configuration file and generate an iso image with cloud-localds. This iso image holds the cloud-init configuration and will be used to setup the system during the bootstrap.
Install cloud-utils
It’s important to NOT install cloud-init on your KVM host machine. This creates a cloud-init service that runs during the boot and tries to reconfigure your host. Something that you probably don’t want on your KVM hypervisor host.
The cloud-util package has all the tool we need to convert the cloud-init configuration files to an iso image.
We’ll create a cloud-init configuration file to update all the packages - which is always a good idea - and to add a user to the system.
A cloud-init configuration file has to start with #cloud-config, remember this is YAML so only use spaces…
We’ll create a password hash that we’ll put into your cloud-init configuration, it’s also possible to use a plain-text password in the configuration with chpasswd or to set the password for the default user. But it’s better to use a hash so nobody can see the password. Keep in mind that is still possible to brute-force the password hash.
Some GNU/Linux distributions have the mkpasswd utility this is not available on centos. The mkpasswd utility is part of the expect package and is something else…
I used a python one-liner to generate the SHA512 password hash
Actually, I got it a couple of months back but I didn’t have time to play with it and it took some time to get some parts from Aliexpress.
The Thinkpad W500 is probably the most powerful system that is compatible with Libreboot, it has a nice high-resolution display with a 1920 x 1200 resolution which is even a higher screen resolution than the Full HD resolution used on most new laptops today.
Binary blobs are bad. Having a closed source binary-only piece of software on your system is not only unacceptable for Free Software activists it also makes it more difficult to review what it really does and makes it more difficult to review it for security concerns.
Having your system vulnerable is also a bad thing of course. Can’t wait to get a computer system with an open CPU architecture like RISC-V.
Preparation
Thinkpad
MAC address
Your MAC address is stored in your BIOS since you’ll overwite the BIOS with Libreboot we need to have the MAC address. Your MAC address is written on Laptop however I recommend to boot from GNU/Linux and copy/paste it from the ifconfig or the ip a command.
EC update
It’s recommended to update your current BIOS to get the latest EC firmware. My system has a CDROM drive I updated the BIOS with a CDROM.
Prepare the Raspberry-pi
It isn’t possible to flash the BIOS with software only on a Lenovo W500/T500, it’s required to put a clip on your BIOS chip and flash the new BIOS with flashrom. I used a Raspberry Pi 1 model B with Raspbian to flash Libreboot .
Enable the SPI port
The SPI port isn’t enabled by default on Raspbian, so we’ll need to enable it.
Open /boot/config.txt in your favorite text editor.
pi@raspberrypi:~ $ sudo apt install git
Reading package lists... Done
Building dependency tree
Reading state information... Done
The following additional packages will be installed:
git-man liberror-perl
Suggested packages:
git-daemon-run | git-daemon-sysvinit git-doc git-el git-email git-gui gitk gitweb git-arch git-cvs
git-mediawiki git-svn
The following NEW packages will be installed:
git git-man liberror-perl
0 upgraded, 3 newly installed, 0 to remove and 0 not upgraded.
Need to get 4,849 kB of archives.
After this operation, 26.4 MB of additional disk space will be used.
Do you want to continue? [Y/n] y
Get:1 http://mirror.nl.leaseweb.net/raspbian/raspbian stretch/main armhf liberror-perl all 0.17024-1 [26.9 kB]
Get:2 http://mirror.nl.leaseweb.net/raspbian/raspbian stretch/main armhf git-man all 1:2.11.0-3+deb9u4 [1,433 kB]
Get:3 http://mirror.nl.leaseweb.net/raspbian/raspbian stretch/main armhf git armhf 1:2.11.0-3+deb9u4 [3,390 kB]
Fetched 4,849 kB in 3s (1,517 kB/s)
Selecting previously unselected package liberror-perl.
(Reading database ... 35178 files and directories currently installed.)
Preparing to unpack .../liberror-perl_0.17024-1_all.deb ...
Unpacking liberror-perl (0.17024-1) ...
Selecting previously unselected package git-man.
Preparing to unpack .../git-man_1%3a2.11.0-3+deb9u4_all.deb ...
Unpacking git-man (1:2.11.0-3+deb9u4) ...
Selecting previously unselected package git.
Preparing to unpack .../git_1%3a2.11.0-3+deb9u4_armhf.deb ...
Unpacking git (1:2.11.0-3+deb9u4) ...
Setting up git-man (1:2.11.0-3+deb9u4) ...
Setting up liberror-perl (0.17024-1) ...
Processing triggers for man-db (2.7.6.1-2) ...
Setting up git (1:2.11.0-3+deb9u4) ...
pi@raspberrypi:~ $
root@raspberrypi:~# apt install flashrom
Reading package lists... Done
Building dependency tree
Reading state information... Done
The following additional packages will be installed:
libftdi1-2 libpci3
The following NEW packages will be installed:
flashrom libftdi1-2 libpci3
0 upgraded, 3 newly installed, 0 to remove and 0 not upgraded.
Need to get 454 kB of archives.
After this operation, 843 kB of additional disk space will be used.
Do you want to continue? [Y/n] y
Get:1 http://mirror.nl.leaseweb.net/raspbian/raspbian stretch/main armhf libpci3 armhf 1:3.5.2-1 [50.9 kB]
Get:2 http://mirror.nl.leaseweb.net/raspbian/raspbian stretch/main armhf libftdi1-2 armhf 1.3-2 [26.8 kB]
Get:3 http://mirror.nl.leaseweb.net/raspbian/raspbian stretch/main armhf flashrom armhf 0.9.9+r1954-1 [377 kB]
Fetched 454 kB in 4s (108 kB/s)
Selecting previously unselected package libpci3:armhf.
(Reading database ... 34656 files and directories currently installed.)
Preparing to unpack .../libpci3_1%3a3.5.2-1_armhf.deb ...
Unpacking libpci3:armhf (1:3.5.2-1) ...
Selecting previously unselected package libftdi1-2:armhf.
Preparing to unpack .../libftdi1-2_1.3-2_armhf.deb ...
Unpacking libftdi1-2:armhf (1.3-2) ...
Selecting previously unselected package flashrom.
Preparing to unpack .../flashrom_0.9.9+r1954-1_armhf.deb ...
Unpacking flashrom (0.9.9+r1954-1) ...
Setting up libftdi1-2:armhf (1.3-2) ...
Processing triggers for libc-bin (2.24-11+deb9u3) ...
Processing triggers for man-db (2.7.6.1-2) ...
Setting up libpci3:armhf (1:3.5.2-1) ...
Setting up flashrom (0.9.9+r1954-1) ...
Processing triggers for libc-bin (2.24-11+deb9u3) ...
root@raspberrypi:~#
Wiring
Wire diagram
It’s useful to get correct flash chip specs, I used a magnifying loupe and a photo camera to get my chip type. After searching the internet I found a very nice blog post from p1trsonhttps://p1trson.blogspot.com/2017/01/journey-to-freedom-part-ii.html about flashing Libreboot on a Thinkpad T400 with the same Flash chip, I used his wiring diagram. Thanks P1trson!
Power off & wiring
Power off your raspberry-pi and wire your flash clip to the raspberry-pi with the above diagram.
1234
root@raspberrypi:~# poweroff
Connection to pi2 closed by remote host.
Connection to pi2 closed.
[staf@vicky ~]$
flashing
test
Test the connection to your flash chip with flashrom. I needed to specify the spispeed=512 to get the connection established.
123456789101112
root@raspberrypi:~# flashrom -p linux_spi:dev=/dev/spidev0.0,spispeed=512
flashrom v0.9.9-r1954 on Linux 4.14.79+ (armv6l)
flashrom is free software, get the source code at https://flashrom.org
Calibrating delay loop... OK.
Found Macronix flash chip "MX25L6405" (8192 kB, SPI) on linux_spi.
Found Macronix flash chip "MX25L6405D" (8192 kB, SPI) on linux_spi.
Found Macronix flash chip "MX25L6406E/MX25L6408E" (8192 kB, SPI) on linux_spi.
Found Macronix flash chip "MX25L6436E/MX25L6445E/MX25L6465E/MX25L6473E" (8192 kB, SPI) on linux_spi.
Multiple flash chip definitions match the detected chip(s): "MX25L6405", "MX25L6405D", "MX25L6406E/MX25L6408E", "MX25L6436E/MX25L6445E/MX25L6465E/MX25L6473E"
Please specify which chip definition to use with the -c <chipname> option.
root@raspberrypi:~#
read old bios
read
Read the original flash twice
1234567891011121314151617
pi@raspberrypi:~ $ sudo flashrom -c "MX25L6405D" -p linux_spi:dev=/dev/spidev0.0,spispeed=512 -r w500bios.rom
flashrom v0.9.9-r1954 on Linux 4.14.79+ (armv6l)
flashrom is free software, get the source code at https://flashrom.org
Calibrating delay loop... OK.
Found Macronix flash chip "MX25L6405D" (8192 kB, SPI) on linux_spi.
Reading flash... done.
pi@raspberrypi:~ $ ls
flashrom test.rom w500bios.rom
pi@raspberrypi:~ $ sudo flashrom -c "MX25L6405D" -p linux_spi:dev=/dev/spidev0.0,spispeed=512 -r w500bios2.rom
flashrom v0.9.9-r1954 on Linux 4.14.79+ (armv6l)
flashrom is free software, get the source code at https://flashrom.org
Calibrating delay loop... OK.
Found Macronix flash chip "MX25L6405D" (8192 kB, SPI) on linux_spi.
Reading flash... done.
pi@raspberrypi:~ $
It always a good idea to verify the gpg signature…
Download the gpg key
1234
pi@raspberrypi:~ $ gpg --recv-keys 0x969A979505E8C5B2
gpg: failed to start the dirmngr '/usr/bin/dirmngr': No such file or directory
gpg: connecting dirmngr at '/run/user/1000/gnupg/S.dirmngr' failed: No such file or directory
gpg: keyserver receive failed: No dirmngr
I needed to install dirmgr separately on my Raspbian installation.
1234567891011121314151617181920
pi@raspberrypi:~ $ sudo apt-get install dirmngr
Reading package lists... Done
Building dependency tree
Reading state information... Done
Suggested packages:
dbus-user-session pinentry-gnome3 tor
The following NEW packages will be installed:
dirmngr
0 upgraded, 1 newly installed, 0 to remove and 0 not upgraded.
Need to get 547 kB of archives.
After this operation, 963 kB of additional disk space will be used.
Get:1 http://mirror.nl.leaseweb.net/raspbian/raspbian stretch/main armhf dirmngr armhf 2.1.18-8~deb9u3 [547 kB]
Fetched 547 kB in 9s (58.3 kB/s)
Selecting previously unselected package dirmngr.
(Reading database ... 36051 files and directories currently installed.)
Preparing to unpack .../dirmngr_2.1.18-8~deb9u3_armhf.deb ...
Unpacking dirmngr (2.1.18-8~deb9u3) ...
Processing triggers for man-db (2.7.6.1-2) ...
Setting up dirmngr (2.1.18-8~deb9u3) ...
pi@raspberrypi:~ $
Try it again…
123456789
pi@raspberrypi:~ $ gpg --recv-keys 0x969A979505E8C5B2
key 969A979505E8C5B2:
1 signature not checked due to a missing key
gpg: /home/pi/.gnupg/trustdb.gpg: trustdb created
gpg: key 969A979505E8C5B2: public key "Leah Rowe (Libreboot signing key) <info@minifree.org>" imported
gpg: no ultimately trusted keys found
gpg: Total number processed: 1
gpg: imported: 1
pi@raspberrypi:~ $
Verify the signature of the checksum file…
123456789
pi@raspberrypi:~ $ gpg --verify SHA512SUMS.sig
gpg: assuming signed data in 'SHA512SUMS'
gpg: Signature made Wed 07 Sep 2016 23:15:17 BST
gpg: using RSA key 969A979505E8C5B2
gpg: Good signature from "Leah Rowe (Libreboot signing key) <info@minifree.org>" [unknown]
gpg: WARNING: This key is not certified with a trusted signature!
gpg: There is no indication that the signature belongs to the owner.
Primary key fingerprint: CDC9 CAE3 2CB4 B7FC 84FD C804 969A 9795 05E8 C5B2
pi@raspberrypi:~ $
Compare the checksum…
1234567
pi@raspberrypi:~/libreboot $ sha512sum libreboot_r20160907_grub_t500_8mb.tar.xz
5325aef526ab6ca359d6613609a4a2345eee47c6d194094553b53996c413431bccdc345838299b347f47bcba8896dd0a6ed3f9b4c88606ead61c3725b580983b libreboot_r20160907_grub_t500_8mb.tar.xz
pi@raspberrypi:~/libreboot $ grep sha512sum 5325aef526ab6ca359d6613609a4a2345eee47c6d194094553b53996c413431bccdc345838299b347f47bcba8896dd0a6ed3f9b4c88606ead61c3725b580983b
grep: 5325aef526ab6ca359d6613609a4a2345eee47c6d194094553b53996c413431bccdc345838299b347f47bcba8896dd0a6ed3f9b4c88606ead61c3725b580983b: No such file or directory
pi@raspberrypi:~/libreboot $ grep 5325aef526ab6ca359d6613609a4a2345eee47c6d194094553b53996c413431bccdc345838299b347f47bcba8896dd0a6ed3f9b4c88606ead61c3725b580983b SHA512SUMS
5325aef526ab6ca359d6613609a4a2345eee47c6d194094553b53996c413431bccdc345838299b347f47bcba8896dd0a6ed3f9b4c88606ead61c3725b580983b ./rom/grub/libreboot_r20160907_grub_t500_8mb.tar.xz
pi@raspberrypi:~/libreboot $
pi@raspberrypi:~/libreboot $ ./ich9gen --macaddress XX:XX:XX:XX:XX:XX
You selected to change the MAC address in the Gbe section. This has been done.
The modified gbe region has also been dumped as src files: mkgbe.c, mkgbe.h
To use these in ich9gen, place them in src/ich9gen/ and re-build ich9gen.
descriptor and gbe successfully written to the file: ich9fdgbe_4m.bin
Now do: dd if=ich9fdgbe_4m.bin of=libreboot.rom bs=1 count=12k conv=notrunc
(in other words, add the modified descriptor+gbe to your ROM image)
descriptor and gbe successfully written to the file: ich9fdgbe_8m.bin
Now do: dd if=ich9fdgbe_8m.bin of=libreboot.rom bs=1 count=12k conv=notrunc
(in other words, add the modified descriptor+gbe to your ROM image)
descriptor and gbe successfully written to the file: ich9fdgbe_16m.bin
Now do: dd if=ich9fdgbe_16m.bin of=libreboot.rom bs=1 count=12k conv=notrunc
(in other words, add the modified descriptor+gbe to your ROM image)
descriptor successfully written to the file: ich9fdnogbe_4m.bin
Now do: dd if=ich9fdnogbe_4m.bin of=yourrom.rom bs=1 count=4k conv=notrunc
(in other words, add the modified descriptor to your ROM image)
descriptor successfully written to the file: ich9fdnogbe_8m.bin
Now do: dd if=ich9fdnogbe_8m.bin of=yourrom.rom bs=1 count=4k conv=notrunc
(in other words, add the modified descriptor to your ROM image)
descriptor successfully written to the file: ich9fdnogbe_16m.bin
Now do: dd if=ich9fdnogbe_16m.bin of=yourrom.rom bs=1 count=4k conv=notrunc
(in other words, add the modified descriptor to your ROM image)
Insert the mac into your rom
1234567
pi@raspberrypi:~/libreboot $ dd if=ich9fdgbe_8m.bin of=libreboot.rom bs=12k count=1 conv=notrunc
1+0 records in
1+0 records out
12288 bytes (12 kB, 12 KiB) copied, 0.00883476 s, 1.4 MB/s
pi@raspberrypi:~/libreboot $ ls -lh libreboot.rom
-rw-r--r-- 1 pi pi 8.0M Jan 27 09:38 libreboot.rom
pi@raspberrypi:~/libreboot $
## flash it
Flash your Libreboot image to your BIOS.
Make sure that you get the Verifying flash... VERIFIED message, if you don’t get this message try it again until you get it. I needed to do it twice…
1234567891011121314151617181920212223
pi@raspberrypi:~/libreboot $ sudo flashrom -c "MX25L6405D" -p linux_spi:dev=/dev/spidev0.0,spispeed=512 -w libreboot.rom
flashrom v0.9.9-r1954 on Linux 4.14.79+ (armv6l)
flashrom is free software, get the source code at https://flashrom.org
Calibrating delay loop... OK.
Found Macronix flash chip "MX25L6405D" (8192 kB, SPI) on linux_spi.
Reading old flash chip contents... done.
Erasing and writing flash chip... Erase/write done.
Verifying flash... FAILED at 0x000c9f01! Expected=0x6b, Found=0xe9, failed byte count from 0x00000000-0x007fffff: 0x2
Your flash chip is in an unknown state.
Please report this on IRC at chat.freenode.net (channel #flashrom) or
mail flashrom@flashrom.org, thanks!
pi@raspberrypi:~/libreboot $
pi@raspberrypi:~/libreboot $ sudo flashrom -c "MX25L6405D" -p linux_spi:dev=/dev/spidev0.0,spispeed=512 -w libreboot.rom
flashrom v0.9.9-r1954 on Linux 4.14.79+ (armv6l)
flashrom is free software, get the source code at https://flashrom.org
Calibrating delay loop... OK.
Found Macronix flash chip "MX25L6405D" (8192 kB, SPI) on linux_spi.
Reading old flash chip contents... done.
Erasing and writing flash chip... Erase/write done.
Verifying flash... VERIFIED.
pi@raspberrypi:~/libreboot $
The intel wifi card that Lenovo uses on the W500 isn’t supported without a binary blob. With the original Lenovo BIOS you are forced to use certified PCI card. Libreboot doesn’t have this restriction this is another advantage of using an alternative BIOS like Libreboot or Coreboot . I replaced wifi an Atheros from ebay
The Intel core duo is still a captable CPU. Even video playback in Full-HD is possible but it takes already a lot of the CPU and the temperature is increasing during the playback.
I installed thinkfan with a more aggresive cooling profile to keep the CPU temperature under control.
Install thinkfan
1234567891011
[staf@snuffel ~]$ yay -S thinkfan
warning: thinkfan-0.9.3-1 is up to date -- reinstalling
resolving dependencies...
looking for conflicting packages...
Packages (1) thinkfan-0.9.3-1
Total Installed Size: 0.11 MiB
Net Upgrade Size: 0.00 MiB
:: Proceed with installation? [Y/n]
I use a Centos 7 virtual system running as a KVM instance with nested KVM virtualasation enabled. The system requiremensts
The minimun requiremenst are:
8 CPU cores
50 GB of free diskspace
8GB RAM
update ….
Make sure that your system is up-to-update
1234567891011121314151617181920
[staf@openstack ~]$ sudo yum update -y
We trust you have received the usual lecture from the local System
Administrator. It usually boils down to these three things:
#1) Respect the privacy of others.
#2) Think before you type.
#3) With great power comes great responsibility.
[sudo] password for staf:
Loaded plugins: fastestmirror
Loading mirror speeds from cached hostfile
* base: distrib-coffee.ipsl.jussieu.fr
* extras: mirror.in2p3.fr
* updates: centos.mirror.fr.planethoster.net
base | 3.6 kB 00:00:00
extras | 3.4 kB 00:00:00
updates | 3.4 kB 00:00:00
No packages marked for update
[staf@openstack ~]$
Install git
We’ll need git to install the ansible playbooks and the Openstack-Ansible installation scripts.
123456789101112
[staf@openstack ~]$ yum install git
Loaded plugins: fastestmirror
You need to be root to perform this command.
[staf@openstack ~]$ sudo yum install git
Loaded plugins: fastestmirror
Loading mirror speeds from cached hostfile
* base: mirror.in2p3.fr
* extras: mirror.in2p3.fr
* updates: centos.mirror.fr.planethoster.net
Package git-1.8.3.1-20.el7.x86_64 already installed and latest version
Nothing to do
[staf@openstack ~]$
Ansible….
This is a bit of a pitfail… The Openstack-Ansible bootstrap script will download and install his own version of ansible and create a link to /usr/local/bin. So /usr/local/bin must be in your $PATH. Ansible shouldn’t be installed on your system or if it is installed it shouln’t be executed instead of the ansible version that is builded with Openstack-Ansible.
On most GNU/Linux distributions have /usr/local/bin and /usr/local/sbin is in the $PATH but not on centos, so we’ll need to add it.
[root@openstack ~]# cd /opt/openstack-ansible
[root@openstack openstack-ansible]#
choose you Openstack releases
Openstack has release shedule about every 6 months the current stable release is Rocky. Every Openstack release has his own branch in the git repo. Each Openstack-Ansible release is tagged in the git repo. So either you’ll need checkout Openstack-Ansible release tag or the bracnh. We’ll checkout the Rocky branch.
[root@openstack openstack-ansible]# git checkout stable/rocky
Branch stable/rocky set up to track remote branch stable/rocky from origin.
Switched to a new branch 'stable/rocky'
[root@openstack openstack-ansible]#
Bootstrap ansible
Execute scripts/bootstrap-ansible.sh this will install the required packages and ansible playbooks.
Stubby is a small dns resolver to encrypt your dns traffic, which makes it perfect to increase end-user privacy. Stubby can be integrated into existing dns setups.
DNSmasq is small dns resolver that can cache dns queries and forward dns traffic to other dns servers.
Unbound is fast validating, caching DNS resolver that supports DNS-over-TLS.
Unbound or dnsmaq are not full feature dns servers like BIND.
The main difference beteen Unbound and DNSmasq is that Unbound can talk the the root servers directly while dnsmasq always needs to forward your dns queries to another dns server - your ISP dns server or a public dns servicve like (Quad9, cloudfare, google, …) -
Unbound has build-in support for DNS-over-TLS. DNSmasq needs an external DNS-over-TLS resolver like Stubby.
Which one to use?
It depends - as always -, Stubby can integrating easily in existing dns setups like dnsmasq. Unbound is one package that does it all and is more feature rich compared to DNSmasq.
OPNsense
I use OPNsense as my firewall. Unbound is the default dns resolver on OPNsense so it makes (OPN)sense to use Unbound.
Choose your upstream DNS service
There’re a few public DNS providers that supports DNS-over-tls the best known are Quad9, cloudfare. Quad9 will block malicious domains on the default dns servers 9.9.9.9/149.112.112.10 while 9.9.9.10 has no security blocklist.
In this article we’ll use Quad9 but you could also with cloudfare or another dns provider that you trust and has support for DNS-over-tls.
Enable DNS-over-TLS
You need to configure your firewall to use your upstream dns provider. You also want to make sure your isp dns servers aren’t used.
Sniffing
If you snif the DNS traffic on your firewall tcpdump -i wan_interface udp port 53 you’ll see that the DNS traffic is unencrypted.
Configuration
To enable DNS-over-TLS we’ll need to reconfigure unbound.
Go to [ Services ] -> [Unbound DNS ] -> [General]
And copy/paste the setting below
to Custom options these settings will reconfigure Unbound to forward the dns for the upstream dns servers Quad9 over ssl.
Verify
If you snif the udp traffic on you firewall with tcpdump -i wan_interface udp port 53 you’ll not see any unencrypted traffic anymore - unless not all your clients are configured to use your firewall as the dns server -.
If your snif TCP PORT 853 tcpdump -i vr1 tcp port 853 we’ll see your encrypted dns-over-tls traffic.
General DNS settings
You also want to make sure that your firewall isn’t configure to use an unecrypted DNS server.
Configuration
Go to [ system ] -> [ settings ] -> [ general ] and set the dns servers also make sure that [ ] Allow DNS server list to be overridden by DHCP/PPP on WAN is unchecked.
Verify
You can verify the configuration by logging on to your firewall over ssh and reviewing the contents of /etc/resolv.conf.
In my previous blog article we install on GNU/Linux which is my main desktop operation system. My NAS and the services that are required to be always running are on FreeBSD.
In this arcticle we will setup Stubby - the DNS Privacy Daemon - on FreeBSD.
Jails
FreeBSD jails are verify nice to keep services separated in a secure way.
ezjail
ezjail is a very nice tool for managing FreeBSD jails.
1
root@rataplan:~ # pkg install ezjail
To loopback or not to loopback….
The loopback ip address is mapped to the jail ip address on FreeBSD by default.
There are two options
use the jail ip address, make sure that you setup a firewall rule if you want disable traffic from external.
use a cloned loopback interface; keep in mind that with a cloned interface this interface is shared between your jails.
We’ll use the jail ip address.
create the jail
We create a new jail and we assign the cloned loop interface with a loopback ip address - this loopback ip address must be unique for each for each jail - and outside interface and ip address.
root@rataplan:/usr/local/etc/ezjail # ezjail-admin start stafdns
Starting jails: stafdns.
/etc/rc.d/jail: WARNING: Per-jail configuration via jail_* variables is obsolete. Please consider migrating to /etc/jail.conf.
root@rataplan:/usr/local/etc/ezjail #
console login and install pkg
Logon to the jail and install pkg we might need to configure a dns server or use a proxy sever to install pkg.
1234567891011121314
root@rataplan:/usr/local/etc/ezjail # ezjail-admin console stafdns
root@stafdns:~ # echo "nameserver 9.9.9.9" > /etc/resolv.conf
root@stafdns:~ # pkg
The package management tool is not yet installed on your system.
Do you want to fetch and install it now? [y/N]: y
Bootstrapping pkg from pkg+http://pkg.FreeBSD.org/FreeBSD:11:amd64/quarterly, please wait...
Verifying signature with trusted certificate pkg.freebsd.org.2013102301... done
[stafdns] Installing pkg-1.10.5_1...
[stafdns] Extracting pkg-1.10.5_1: 100%
pkg: not enough arguments
Usage: pkg [-v] [-d] [-l] [-N] [-j <jail name or id>|-c <chroot path>|-r <rootdir>] [-C <configuration file>] [-R <repo config dir>] [-o var=value] [-4|-6] <command> [<args>]
For more information on available commands and options see 'pkg help'.
root@stafdns:~ #
Install stubby
Stubby available in the FreeBSD Ports in the getdns package, …but it isn’t installed when you install the binary package. To install stubby we need to it from source.
dig
To debug dns issues dig a handy tool to have….
12345678910111213141516171819
root@rataplan:/usr/ports/dns/getdns # pkg install bind-tools
Updating FreeBSD repository catalogue...
FreeBSD repository is up to date.
All repositories are up to date.
Updating database digests format: 100%
The following 4 package(s) will be affected (of 0 checked):
New packages to be INSTALLED:
bind-tools: 9.12.2P1
idnkit: 1.0_7
py27-ply: 3.11
json-c: 0.13
Number of packages to be installed: 4
The process will require 42 MiB more space.
4 MiB to be downloaded.
Proceed with this action? [y/N]: y
Update your ports tree
Physical system
On a physical FreeBSD system execute portsnap fetch and portsnap extract
123456789101112131415
root@rataplan:~ # portsnap fetch
Looking up portsnap.FreeBSD.org mirrors... 6 mirrors found.
Fetching snapshot tag from ec2-eu-west-1.portsnap.freebsd.org... done.
Fetching snapshot metadata... done.
Updating from Sat Sep 8 09:31:35 CEST 2018 to Sun Sep 9 09:51:49 CEST 2018.
Fetching 4 metadata patches... done.
Applying metadata patches... done.
Fetching 0 metadata files... done.
Fetching 44 patches.
(44/44) 100.00% done.
done.
Applying patches...
done.
Fetching 2 new ports or files... done.
root@rataplan:~ #
12345678910111213141516171819202122232425
root@rataplan:~ # portsnap extract
/usr/ports/.arcconfig
/usr/ports/.gitattributes
/usr/ports/.gitauthors
/usr/ports/.gitignore
/usr/ports/.gitmessage
/usr/ports/CHANGES
/usr/ports/CONTRIBUTING.md
/usr/ports/COPYRIGHT
/usr/ports/GIDs
/usr/ports/Keywords/desktop-file-utils.ucl
/usr/ports/Keywords/fc.ucl
/usr/ports/Keywords/fcfontsdir.ucl
<snip>
/usr/ports/x11/xzoom/
/usr/ports/x11/yad/
/usr/ports/x11/yakuake-kde4/
/usr/ports/x11/yakuake/
/usr/ports/x11/yalias/
/usr/ports/x11/yeahconsole/
/usr/ports/x11/yelp/
/usr/ports/x11/zenity/
Building new INDEX files... done.
Jail
I use ezjail to manage my FreeBSD jails. Execute the ezjail-admin update -P to update the ports tree inside your jails.
root@stafproxy:/usr/ports/dns/getdns # make
===> License BSD3CLAUSE accepted by the user
===> getdns-1.4.2 depends on file: /usr/local/sbin/pkg - found
=> getdns-1.4.2.tar.gz doesn't seem to exist in /var/ports/distfiles/.
=> Attempting to fetch https://getdnsapi.net/dist/getdns-1.4.2.tar.gz
getdns-1.4.2.tar.gz 100% of 1034 kB 1092 kBps 00m01s
===> Fetching all distfiles required by getdns-1.4.2 for building
===> Extracting for getdns-1.4.2
=> SHA256 Checksum OK for getdns-1.4.2.tar.gz.
<snip>
/usr/bin/install -c -m 644 getdns_service_sync.3 /var/ports/basejail/usr/ports/dns/getdns/work/stage/usr/local/man/man3
/usr/bin/install -c -m 644 getdns_validate_dnssec.3 /var/ports/basejail/usr/ports/dns/getdns/work/stage/usr/local/man/man3
/usr/bin/strip /var/ports/basejail/usr/ports/dns/getdns/work/stage/usr/local/lib/libgetdns*.so.*
/usr/bin/strip /var/ports/basejail/usr/ports/dns/getdns/work/stage/usr/local/bin/getdns_*
/usr/bin/strip /var/ports/basejail/usr/ports/dns/getdns/work/stage/usr/local/bin/stubby
/bin/mv /var/ports/basejail/usr/ports/dns/getdns/work/stage/usr/local/etc/stubby/stubby.yml /var/ports/basejail/usr/ports/dns/getdns/work/stage/usr/local/etc/stubby/stubby.yml.sample
====> Compressing man pages (compress-man)
===> Staging rc.d startup script(s)
root@stafproxy:/usr/ports/dns/getdns # make install
===> Installing for getdns-1.4.2
===> Checking if getdns already installed
===> Registering installation for getdns-1.4.2
[stafproxy] Installing getdns-1.4.2...
***
*** !!! IMPORTANT !!!! libgetdns needs a DNSSEC trust anchor!
***
*** For the library to be able to perform DNSSEC, the root
*** trust anchor needs to be present in presentation format
*** in the file:
*** /usr/local/etc/unbound/root.key
***
*** We recomend using unbound-anchor to retrieve and install
*** the root trust anchor like this:
*** su -m unbound -c /usr/local/sbin/unbound-anchor
***
===> SECURITY REPORT:
This port has installed the following files which may act as network
servers and may therefore pose a remote security risk to the system.
/usr/local/lib/libgetdns.a(stub.o)
/usr/local/lib/libgetdns.so.10.0.2
/usr/local/lib/libgetdns.a(server.o)
This port has installed the following startup scripts which may cause
these network services to be started at boot time.
/usr/local/etc/rc.d/stubby
If there are vulnerabilities in these programs there may be a security
risk to the system. FreeBSD makes no guarantee about the security of
ports included in the Ports Collection. Please type 'make deinstall'
to deinstall the port if this is a concern.
For more information, and contact details about the security
status of this software, see the following webpage:
https://getdnsapi.net/
root@stafproxy:/usr/ports/dns/getdns #
Lock the package to avoid that the package gets replaced by a getdns package without stubby.
1234
root@stafproxy:/usr/ports/dns/getdns # pkg lock getdns
getdns-1.4.2: lock this package? [y/N]: y
Locking getdns-1.4.2
root@stafproxy:/usr/ports/dns/getdns #
Configure stubby
Enable the stubby service
Use sysrc to enable the stubby service…
1234567891011
root@stafproxy:/usr/local/etc # service stubby start
Cannot 'start' stubby. Set stubby_enable to YES in /etc/rc.conf or use 'onestart' instead of 'start'.
root@stafproxy:/usr/local/etc # service stubby rcvar
# stubby
#
stubby_enable="NO"
# (default: "")
root@stafproxy:/usr/local/etc # sysrc stubby_enable="YES"
stubby_enable: -> YES
root@stafproxy:/usr/local/etc #
choose your upstream dns provider
Edit the stubby.yml file and uncomment the upstream dns server that you want the use. Stubby will loadbalance the dns traffic to all configured upstream dns servers by default. This is configured with the round_robin_upstreams directive, if set to 1 the traffic is loadbalanced, if set 0 stubby will use the first configured dns server.
1
root@rataplan:/usr/local/etc # vi stubby/stubby.yml
Change the port
We’ll setup dnsmasq to cache our dns requests modify the listen_addresses directive and set the port 53000
123
listen_addresses:
- 127.0.0.1@53000
- 0::1@53000
Start it
1234
root@stafproxy:/usr/local/etc # service stubby start
Starting stubby.
[07:51:37.865826] STUBBY: Read config from file /usr/local/etc/stubby/stubby.yml
root@stafproxy:/usr/local/etc #
root@stafproxy:/root # pkg install dnsmasq
Updating FreeBSD repository catalogue...
FreeBSD repository is up to date.
All repositories are up to date.
The following 1 package(s) will be affected (of 0 checked):
New packages to be INSTALLED:
dnsmasq: 2.79,1
Number of packages to be installed: 1
329 KiB to be downloaded.
Proceed with this action? [y/N]: y
[stafproxy] [1/1] Fetching dnsmasq-2.79,1.txz: 100% 329 KiB 336.4kB/s 00:01
Checking integrity... done (0 conflicting)
[stafproxy] [1/1] Installing dnsmasq-2.79,1...
[stafproxy] [1/1] Extracting dnsmasq-2.79,1: 100%
Message from dnsmasq-2.79,1:
*** To enable dnsmasq, edit /usr/local/etc/dnsmasq.conf and
*** set dnsmasq_enable="YES" in /etc/rc.conf[.local]
***
*** Further options and actions are documented inside
*** /usr/local/etc/rc.d/dnsmasq
root@stafproxy:/root #
