DoT and DoH (encrypted DNS)


It is important to understand limits of encrypting only DNS traffic. Relevant security analysis can be found in article Simran Patil and Nikita Borisov. 2019. What can you learn from an IP? See slides or the article itself.

DoT and DoH encrypt DNS traffic with Transport Layer Security (TLS) protocol and thus protects DNS traffic from certain types of attacks.

DNS-over-TLS (DoT)

DNS-over-TLS server (RFC 7858) can be configured using tls kind in net.listen(). It is enabled on localhost by default.

For certificate configuration, refer to Configuration options.

DNS-over-HTTPS (DoH)


Knot Resolver currently offers two DoH implementations. It is recommended to use this new implementation, which is more reliable, scalable and has fewer dependencies. Make sure to use doh2 kind in net.listen() to select this implementation.


Independent information about political controversies around the DoH deployment by default can be found in blog posts DNS Privacy at IETF 104 and More DOH by Geoff Huston and Centralised DoH is bad for Privacy, in 2019 and beyond by Bert Hubert.

DNS-over-HTTPS server (RFC 8484) can be configured using doh2 kind in net.listen().

This implementation supports HTTP/2 (RFC 7540). Queries can be sent to the /dns-query endpoint, e.g.:

$ kdig @ +https AAAA

Only TLS version 1.3 (or higher) is supported with DNS-over-HTTPS. The additional considerations for TLS 1.2 required by HTTP/2 are not implemented (RFC 7540#section-9.2).


Take care when configuring your server to listen on well known HTTPS port. If an unrelated HTTPS service is running on the same port with REUSEPORT enabled, you will end up with both services malfunctioning.

Configuration options


These settings affect both DNS-over-TLS and DNS-over-HTTPS (except the legacy implementation).

A self-signed certificate is generated by default. For serious deployments it is strongly recommended to configure your own TLS certificates signed by a trusted CA. This is done using function net.tls().

net.tls([cert_path][, key_path])

Get/set path to a server TLS certificate and private key for DoT and DoH.

Example output:

> net.tls("/etc/knot-resolver/server-cert.pem", "/etc/knot-resolver/server-key.pem")
> net.tls()  -- print configured paths
("/etc/knot-resolver/server-cert.pem", "/etc/knot-resolver/server-key.pem")
net.tls_sticket_secret([string with pre-shared secret])

Set secret for TLS session resumption via tickets, by RFC 5077.

The server-side key is rotated roughly once per hour. By default or if called without secret, the key is random. That is good for long-term forward secrecy, but multiple kresd instances won’t be able to resume each other’s sessions.

If you provide the same secret to multiple instances, they will be able to resume each other’s sessions without any further communication between them. This synchronization works only among instances having the same endianess and time_t structure and size (sizeof(time_t)).

For good security the secret must have enough entropy to be hard to guess, and it should still be occasionally rotated manually and securely forgotten, to reduce the scope of privacy leak in case the secret leaks eventually.


Setting the secret is probably too risky with TLS <= 1.2. GnuTLS stable release supports TLS 1.3 since 3.6.3 (summer 2018). Therefore setting the secrets should be considered experimental for now and might not be available on your system.

net.tls_sticket_secret_file([string with path to a file containing pre-shared secret])

The same as net.tls_sticket_secret(), except the secret is read from a (binary) file.

net.tls_padding([true | false])

Get/set EDNS(0) padding of answers to queries that arrive over TLS transport. If set to true (the default), it will use a sensible default padding scheme, as implemented by libknot if available at compile time. If set to a numeric value >= 2 it will pad the answers to nearest padding boundary, e.g. if set to 64, the answer will have size of a multiple of 64 (64, 128, 192, …). If set to false (or a number < 2), it will disable padding entirely.