Secure (DNS-over-TLS) Adblocking (Pi-hole) Recursive (unbound) Server System setup.
- Open Source DNS system
- Pi-hole based adblocking
- Recursive DNS (see this link here)
- DNS-over-TLS support (specifically for Android)
- Optional DNS-over-HTTPS support (dnsdist allows support for both)
- Upload to Github and promote solution for use
- Requirements list
- Dependencies list
- Create install script for Ubuntu and Debian-based distros
- Post initial code with README on Github
- Promote code and ask for contributions or feature requests
- Create easy-to-use install scripts
- Pi-Hole with https
- Let's Encrypt
- Unbound DNS
- DNS-over-TLS support using Nginx transparent proxy, stunnel4, or dnsdist
- Firewall using ufw
- Pi-hole
- Unbound
- Nginx, stunnel4, or dnsdist
- ufw firewall
- Certbot for Let's Encrypt or ZeroSSL
- Let's Encrypt or ZeroSSL certificate(s)
Please see Contributing page for more info.
Note: Debian-based installer for Ubuntu 20.04 or 20.10
click to show installation instructions
sudo apt update
sudo apt -y install unbound ufw software-properties-common dns-root-data
Unbound causes an issue with dnsmasq until it is set up properly
sudo systemctl disable unbound
sudo systemctl stop unbound
curl -sSL https://install.pi-hole.net | bash
- Take default settings unless you would like to tweak it to your preferences.
- Select standard settings and any upstream DNS servers of choice. These settings will be changed later
- Select any block lists of choice. These can be modified later. By default I select all block lists available in the installer
- Make sure to install the web interface. This is useful in changing settings later.
pihole -a -p
Note: The following is optional only if https is desired for the Web Interface
click to show installation instructions
Information can be found here and here
sudo snap install core; sudo snap refresh core
sudo snap install --classic certbot
sudo ln -s /snap/bin/certbot /usr/bin/certbot
sudo service lighttpd stop
Obtain the certificate interactively
sudo certbot certonly
Follow the documentation here to enable https for lighttpd here
sudo service lighttpd start
click to show installation instructions
Note: This will overwrite your Nginx config, but will back up the current config to /etc/nginx/nginx.conf
.
- Run the following command:
curl -s https://raw.githubusercontent.com/clayauld/Secure-Adblocking-DNS/master/nginx/install.sh | sudo bash
-
Open
/etc/nginx/streams/DoT_dns
and edit the location of your certs to the correct location(s) for your server. -
Run the following command to reload the Nginx config.
sudo systemctl restart nginx
click to show installation instructions
Note: Setup documentation found here
The file should have the following contents:
sslVersion = TLSv1.2
chroot = /var/run/stunnel4
setuid = stunnel4
setgid = stunnel4
pid = /stunnel.pid
[dns]
cert = /etc/letsencrypt/live/example.domain.com/fullchain.pem
key = /etc/letsencrypt/live/example.domain.com/privkey.pem
accept = 853
connect = 127.0.0.1:53
#TIMEOUTidle = 1
#TIMEOUTclose = 1
#TIMEOUTbusy = 1
Enabled=1
sudo systemctl enable stunnel4
sudo systemctl start stunnel4
sudo systemctl status stunnel4
The output should show the service running and clients connecting
click to show installation instructions
Note: Setup documentation found at here
sudo apt install dnsdist
The file should have the following contents:
- Note: change
dns.example.com
to your correct hostname
addACL('0.0.0.0/0')
-- path for certs and listen address for DoT ipv4,
-- by default listens on port 853.
-- Set X(int) for tcp fast open queue size.
addTLSLocal("0.0.0.0", "/etc/letsencrypt/live/dns.example.com/fullchain.pem", "/etc/letsencrypt/live/dns.example.com/privkey.pem", { doTCP=true, reusePort=true, tcpFastOpenSize=64 })
-- path for certs and listen address for DoH ipv4,
-- by default listens on port 443.
-- Set X(int) for tcp fast open queue size.
--
-- In this example we listen directly on port 443. However, since the DoH queries are simple HTTPS requests, the server can be hidden behind Nginx or Haproxy.
addDOHLocal("0.0.0.0", "/etc/letsencrypt/live/dns.example.com/fullchain.pem", "/etc/letsencrypt/live/dns.example.com/privkey.pem", "/dns-query", { doTCP=true, reusePort=true, tcpFastOpenSize=64 })
-- set X(int) number of queries to be allowed per second from a IP
addAction(MaxQPSIPRule(50), DropAction())
-- drop ANY queries sent over udp
addAction(AndRule({QTypeRule(DNSQType.ANY), TCPRule(false)}), DropAction())
-- set X number of entries to be in dnsdist cache by default
-- memory will be preallocated based on the X number
pc = newPacketCache(10000, {maxTTL=86400})
getPool(""):setCache(pc)
-- server policy to choose the downstream servers for recursion
setServerPolicy(leastOutstanding)
-- Here we define our backend, the pihole dns server
newServer({address="127.0.0.1:53", name="127.0.0.1:53", checkInterval=3600})
setMaxTCPConnectionsPerClient(1000) -- set X(int) for number of tcp connections from a single client. Useful for rate limiting the concurrent connections.
setMaxTCPQueriesPerConnection(100) -- set X(int) , similiar to addAction(MaxQPSIPRule(X), DropAction())
sudo systemctl enable dnsdist
sudo systemctl start dnsdist
sudo systemctl status dnsdist
The output should show the service running and clients connecting
click to show installation instructions
Note: This set up was derived from this site.
sudo nano /etc/unbound/unbound.conf.d/pi-hole.conf
Put the following configurations into pi-hole.conf (source https://github.com/anudeepND/pihole-unbound)
server:
# The verbosity number, level 0 means no verbosity, only errors.
# Level 1 gives operational information. Level 2 gives detailed
# operational information. Level 3 gives query level information,
# output per query. Level 4 gives algorithm level information.
# Level 5 logs client identification for cache misses. Default is
# level 1.
verbosity: 0
interface: 127.0.0.1
port: 5335
do-ip4: yes
do-udp: yes
do-tcp: yes
# May be set to yes if you have IPv6 connectivity
do-ip6: yes
# You want to leave this to no unless you have *native* IPv6. With 6to4 and
# Terredo tunnels your web browser should favor IPv4 for the same reasons
prefer-ip6: no
# control which client ips are allowed to make (recursive) queries to this
# server. Specify classless netblocks with /size and action. By default
# everything is refused, except for localhost. Choose deny (drop message),
# refuse (polite error reply), allow (recursive ok), allow_snoop (recursive
# and nonrecursive ok)
access-control: 127.0.0.0/8 allow
#access-control: 10.8.0.0/24 allow
#access-control: 10.16.0.0/24 allow
#access-control: 192.168.6.0/24 allow
#access-control: 192.168.8.0/24 allow
# Use this only when you downloaded the list of primary root servers!
# Read the root hints from this file. Make sure to
# update root.hints evry 5-6 months.
root-hints: "/usr/share/dns/root.hints"
# Trust glue only if it is within the servers authority
harden-glue: yes
# Ignore very large queries.
harden-large-queries: yes
# Require DNSSEC data for trust-anchored zones, if such data is absent, the zone becomes BOGUS
# If you want to disable DNSSEC, set harden-dnssec stripped: no
harden-dnssec-stripped: yes
# Number of bytes size to advertise as the EDNS reassembly buffer
# size. This is the value put into datagrams over UDP towards
# peers. The actual buffer size is determined by msg-buffer-size
# (both for TCP and UDP).
edns-buffer-size: 1232
# Rotates RRSet order in response (the pseudo-random
# number is taken from Ensure privacy of local IP
# ranges the query ID, for speed and thread safety).
# private-address: 192.168.0.0/16
rrset-roundrobin: yes
# Time to live minimum for RRsets and messages in the cache. If the minimum
# kicks in, the data is cached for longer than the domain owner intended,
# and thus less queries are made to look up the data. Zero makes sure the
# data in the cache is as the domain owner intended, higher values,
# especially more than an hour or so, can lead to trouble as the data in
# the cache does not match up with the actual data anymore
cache-min-ttl: 300
cache-max-ttl: 86400
# Have unbound attempt to serve old responses from cache with a TTL of 0 in
# the response without waiting for the actual resolution to finish. The
# actual resolution answer ends up in the cache later on.
serve-expired: yes
# Harden against algorithm downgrade when multiple algorithms are
# advertised in the DS record.
harden-algo-downgrade: yes
# Ignore very small EDNS buffer sizes from queries.
harden-short-bufsize: yes
# Refuse id.server and hostname.bind queries
hide-identity: yes
# Report this identity rather than the hostname of the server.
identity: "Server"
# Refuse version.server and version.bind queries
hide-version: yes
# Prevent the unbound server from forking into the background as a daemon
do-daemonize: no
# Number of bytes size of the aggressive negative cache.
neg-cache-size: 4M
# Send minimum amount of information to upstream servers to enhance privacy
qname-minimisation: yes
# Deny queries of type ANY with an empty response.
# Works only on version 1.8 and above
deny-any: yes
# Do no insert authority/additional sections into response messages when
# those sections are not required. This reduces response size
# significantly, and may avoid TCP fallback for some responses. This may
# cause a slight speedup
minimal-responses: yes
# Perform prefetching of close to expired message cache entries
# This only applies to domains that have been frequently queried
# This flag updates the cached domains
prefetch: yes
# Fetch the DNSKEYs earlier in the validation process, when a DS record is
# encountered. This lowers the latency of requests at the expense of little
# more CPU usage.
prefetch-key: yes
# One thread should be sufficient, can be increased on beefy machines. In reality for
# most users running on small networks or on a single machine, it should be unnecessary
# to seek performance enhancement by increasing num-threads above 1.
num-threads: 1
# more cache memory. rrset-cache-size should twice what msg-cache-size is.
msg-cache-size: 50m
rrset-cache-size: 100m
# Faster UDP with multithreading (only on Linux).
so-reuseport: yes
# Ensure kernel buffer is large enough to not lose messages in traffix spikes
so-rcvbuf: 4m
so-sndbuf: 4m
# Set the total number of unwanted replies to keep track of in every thread.
# When it reaches the threshold, a defensive action of clearing the rrset
# and message caches is taken, hopefully flushing away any poison.
# Unbound suggests a value of 10 million.
unwanted-reply-threshold: 100000
# Minimize logs
# Do not print one line per query to the log
log-queries: no
# Do not print one line per reply to the log
log-replies: no
# Do not print log lines that say why queries return SERVFAIL to clients
log-servfail: no
# Do not print log lines to inform about local zone actions
log-local-actions: no
# If no logfile is specified, syslog is used
logfile: "/var/log/unbound/unbound.log"
# Ensure privacy of local IP ranges
private-address: 192.168.0.0/16
private-address: 169.254.0.0/16
private-address: 172.16.0.0/12
private-address: 10.0.0.0/8
private-address: fd00::/8
private-address: fe80::/10
# Allow the domain (and its subdomains) to contain private addresses.
# local-data statements are allowed to contain private addresses too.
#private-domain: "localnetwork.local"
# locally served zones can be configured for the machines on the LAN.
#local-zone: "localnetwork.local" static
Note: This is optional.
unbound-checkconf /etc/unbound/unbound.conf.d/pi-hole.conf
sudo systemctl enable unbound
sudo systemctl start unbound
sudo systemctl status unbound
The first query will be slow but the subsequent queries will resolve under 1ms.
dig github.com @127.0.0.1 -p 5335
The first command should give a status report of SERVFAIL and no IP address. The second should give NOERROR plus an IP address.
dig sigfail.verteiltesysteme.net @127.0.0.1 -p 5335
dig sigok.verteiltesysteme.net @127.0.0.1 -p 5335
Note: Add this next part if you want to enable HTTPS connection to Pi-Hole Web UI
click to show installation instructions
sudo apt install nginx
sudo nano /etc/nginx/sites-available/pihole-redirect
The file should have the following contents:
server {
listen 444 ssl;
server_name example.domain.com;
#SSL
ssl_certificate /etc/letsencrypt/live/example.domain.com/fullchain.pem;
ssl_certificate_key /etc/letsencrypt/live/example.domain.com/privkey.pem;
# Deny access to root
location / {
autoindex off;
deny all;
}
location /admin {
rewrite /(.*) /$1 break;
proxy_pass http://127.0.0.1:80/admin/;
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
proxy_read_timeout 90;
}
}
sudo rm -rf /etc/nginx/sites-enabled/*
sudo ln -s /etc/nginx/sites-available/pihole-redirect /etc/nginx/sites-enabled/pihole-redirect
sudo service nginx restart
click to show installation instructions
In order to experience high speed and low latency DNS resolution, you need to make some changes to your Pi-hole. These configurations are crucial because if you skip these steps you may experience very slow response times:
- Open the configuration file
/etc/pihole/setupVars.conf
and add/modify theCACHE_SIZE
entry, shown below.
CACHE_SIZE=0
- Run
pihole -r
and select the "Repair" option. This will set the cache size to zero in/etc/dnsmasq.d/01-pihole.conf
.
- This step is important because the caching is already handled by the Unbound. Please note that if the changes are only made to the file
/etc/dnsmasq.d/01-pihole.conf
they will be overwritten once you update/modify Pi-hole.
- When you're using unbound you're relying on that for DNSSEC validation and caching. Setting Pi-hole to do this same things is just going to waste time validating DNSSEC twice. In order to resolve this issue you need to untick the
Use DNSSEC
option in Pi-hole web interface by navigating toSettings > DNS > Advanced DNS settings
.
click to show installation instructions
Next steps to set Pi-hole's upstream DNS server to the unbound service
- Log into Pi-hole web interface
- Go to Settings --> DNS
- Set the Upstream DNS Servers to the locahost instance of unbound and deselect all other upstream servers
- Reboot Server to Apply all changes and check configuration
I would like to thank rajannpatel for posting his repo about this topic. His guide helped contribute to my current setup and notes. See his guide at here.](https://github.com/rajannpatel/Pi-Hole-PiVPN-on-Google-Compute-Engine-Free-Tier-with-Full-Tunnel-and-Split-Tunnel-OpenVPN-Configs)
AnudeepND has also provided some great sources of information. See this link here.