RESOLVED.CONF(5)

✨ NAME ✨

resolved.conf, resolved.conf.d — Network Name Resolution configuration files. Bestie, this is the resolved.conf(5) page, and it's literally about your local DNS + LLMNR name resolution config. 💖

🌸 SYNOPSIS 🌸

/etc/systemd/resolved.conf
/run/systemd/resolved.conf
/usr/local/lib/systemd/resolved.conf
/usr/lib/systemd/resolved.conf
/etc/systemd/resolved.conf.d/*.conf
/run/systemd/resolved.conf.d/*.conf
/usr/local/lib/systemd/resolved.conf.d/*.conf
/usr/lib/systemd/resolved.conf.d/*.conf

🎀 DESCRIPTION 🎀

These configuration files control local DNS and LLMNR name resolution. That's the whole assignment. 🫶

💅 CONFIGURATION DIRECTORIES AND PRECEDENCE 💅

The default configuration is set during compilation, so configuration is only needed when it is necessary to deviate from those defaults. The main configuration file is loaded from one of the listed directories in order of priority, and only the first file found is used:

• /etc/systemd/
• /run/systemd/
• /usr/local/lib/systemd/ [1]
• /usr/lib/systemd/

The vendor version of the file contains commented out entries showing the defaults as a guide to the administrator. Local overrides can also be created by creating drop-ins, as described below. The main configuration file can also be edited for this purpose (or a copy in /etc/ if it is shipped under /usr/), however using drop-ins for local configuration is recommended over modifications to the main configuration file.

In addition to the main configuration file, drop-in configuration snippets are read from /usr/lib/systemd/*.conf.d/, /usr/local/lib/systemd/*.conf.d/, and /etc/systemd/*.conf.d/. Those drop-ins have higher precedence and override the main configuration file. Files in the *.conf.d/ configuration subdirectories are sorted by their filename in lexicographic order, regardless of in which of the subdirectories they reside. When multiple files specify the same option, for options which accept just a single value, the entry in the file sorted last takes precedence, and for options which accept a list of values, entries are collected as they occur in the sorted files.

When packages need to customize the configuration, they can install drop-ins under /usr/. Files in /etc/ are reserved for the local administrator, who may use this logic to override the configuration files installed by vendor packages. Drop-ins have to be used to override package drop-ins, since the main configuration file has lower precedence. It is recommended to prefix all filenames in those subdirectories with a two-digit number and a dash, to simplify the ordering. This also defines a concept of drop-in priorities to allow OS vendors to ship drop-ins within a specific range lower than the range used by users. This should lower the risk of package drop-ins accidentally overriding drop-ins defined by users. It is recommended to use the range 10-40 for drop-ins in /usr/ and the range 60-90 for drop-ins in /etc/ and /run/, to make sure that local and transient drop-ins take priority over drop-ins shipped by the OS vendor.

To disable a configuration file supplied by the vendor, the recommended way is to place a symlink to /dev/null in the configuration directory in /etc/, with the same filename as the vendor configuration file.

🔥 OPTIONS 🔥

The following options are available in the [Resolve] section:

• DNS= — A space-separated list of IPv4 and IPv6 addresses to use as system DNS servers. Each address can optionally take a port number separated with :, a network interface name or index separated with %, and a Server Name Indication (SNI) separated with #. When an IPv6 address is specified with a port number, then the address must be in the square brackets. That is, the acceptable full formats are 111.222.333.444:9953%ifname#example.com for IPv4 and [1111:2222::3333]:9953%ifname#example.com for IPv6. DNS requests are sent to one of the listed DNS servers in parallel to suitable per-link DNS servers acquired from systemd-networkd.service(8) or set at runtime by external applications. For compatibility reasons, if this setting is not specified, the DNS servers listed in /etc/resolv.conf are used instead, if that file exists and any servers are configured in it. This setting defaults to the empty list.

Added in version 213.

• FallbackDNS= — A space-separated list of IPv4 and IPv6 addresses to use as the fallback DNS servers. Please see DNS= for acceptable format of addresses. Any per-link DNS servers obtained from systemd-networkd.service(8) take precedence over this setting, as do any servers set via DNS= above or /etc/resolv.conf. This setting is hence only used if no other DNS server information is known. If this option is not given, a compiled-in list of DNS servers is used instead.

Added in version 216.

• Domains= — A space-separated list of domains, optionally prefixed with ~, used for two distinct purposes described below. Defaults to the empty list.

Any domains not prefixed with ~ are used as search suffixes when resolving single-label hostnames (domain names which contain no dot), in order to qualify them into fully-qualified domain names (FQDNs). These "search domains" are strictly processed in the order they are specified in, until the name with the suffix appended is found. For compatibility reasons, if this setting is not specified, the search domains listed in /etc/resolv.conf with the search keyword are used instead, if that file exists and any domains are configured in it.

The domains prefixed with ~ are called "route-only domains". All domains listed here (both search domains and route-only domains after removing the ~ prefix) define a search path that preferably directs DNS queries to this interface. This search path has an effect only when suitable per-link DNS servers are known. Such servers may be defined through the DNS= setting (see above) and dynamically at run time, for example from DHCP leases. If no per-link DNS servers are known, route-only domains have no effect.

Use the construct ~. (which is composed from ~ to indicate a route-only domain and . to indicate the DNS root domain that is the implied suffix of all DNS domains) to use the DNS servers defined for this link preferably for all domains.

See "Protocols and Routing" in systemd-resolved.service(8) for details of how search and route-only domains are used.

Note that configuring the MulticastDNS domain local as search or routing domain has the effect of routing lookups for this domain to classic unicast DNS. This may be used to provide compatibility with legacy installations that use this domain in a unicast DNS context, against the IANA assignment of this domain to pure MulticastDNS purposes. Search and routing domains are a unicast DNS concept, they cannot be used to resolve single-label lookups via MulticastDNS.

Added in version 229.

• LLMNR= — Takes a boolean argument or resolve. Controls Link-Local Multicast Name Resolution support (RFC 4795[2]) on the local host. If true, enables full LLMNR responder and resolver support. If false, disables both. If set to resolve, only resolution support is enabled, but responding is disabled. Note that systemd-networkd.service(8) also maintains per-link LLMNR settings. LLMNR will be enabled on a link only if the per-link and the global setting is on.

Added in version 216.

• MulticastDNS= — Takes a boolean argument or resolve. Controls Multicast DNS support (RFC 6762[3]) on the local host. If true, enables full Multicast DNS responder and resolver support. If false, disables both. If set to resolve, only resolution support is enabled, but responding is disabled. Note that systemd-networkd.service(8) also maintains per-link Multicast DNS settings. Multicast DNS will be enabled on a link only if the per-link and the global setting is on.

Added in version 234.

• DNSSEC= — Takes a boolean argument or allow-downgrade.

If set to true, all DNS lookups are DNSSEC-validated locally (excluding LLMNR and Multicast DNS). If the response to a lookup request is detected to be invalid a lookup failure is returned to applications. Note that this mode requires a DNS server that supports DNSSEC. If the DNS server does not properly support DNSSEC all validations will fail.

If set to allow-downgrade, DNSSEC validation is attempted, but if the server does not support DNSSEC properly, DNSSEC mode is automatically disabled. Note that this mode makes DNSSEC validation vulnerable to "downgrade" attacks, where an attacker might be able to trigger a downgrade to non-DNSSEC mode by synthesizing a DNS response that suggests DNSSEC was not supported.

If set to false, DNS lookups are not DNSSEC validated.

Note that DNSSEC validation requires retrieval of additional DNS data, and thus results in a small DNS lookup time penalty.

DNSSEC requires knowledge of "trust anchors" to prove data integrity. The trust anchor for the Internet root domain is built into the resolver, additional trust anchors may be defined with dnssec-trust-anchors.d(5). Trust anchors may change at regular intervals, and old trust anchors may be revoked. In such a case DNSSEC validation is not possible until new trust anchors are configured locally or the resolver software package is updated with the new root trust anchor. In effect, when the built-in trust anchor is revoked and DNSSEC= is true, all further lookups will fail, as it cannot be proved anymore whether lookups are correctly signed, or validly unsigned. If DNSSEC= is set to allow-downgrade the resolver will automatically turn off DNSSEC validation in such a case.

Client programs looking up DNS data will be informed whether lookups could be verified using DNSSEC, or whether the returned data could not be verified (either because the data was found unsigned in the DNS, or the DNS server did not support DNSSEC or no appropriate trust anchors were known). In the latter case, it is assumed that client programs employ a secondary scheme to validate the returned DNS data, should this be required.

It is recommended to set DNSSEC= to true on systems where it is known that the DNS server supports DNSSEC correctly, and where software or trust anchor updates happen regularly. On other systems it is recommended to set DNSSEC= to allow-downgrade.

Ok so real talk, bestie: that global DNSSEC setting isn't the only one in the game 💅 systemd-networkd.service(8) also keeps its own per-link DNSSEC settings. For system DNS servers (see above, babe), only the global DNSSEC setting is in effect. For per-link DNS servers the per-link setting is what runs the show — unless it's unset, in which case the global setting steps in instead. Iconic delegation behavior, honestly.

Site-private DNS zones generally conflict with DNSSEC operation 😭 — unless a negative trust anchor (if the private zone isn't signed) or a positive trust anchor (if the private zone is signed) is configured for them. And if "allow-downgrade" mode is selected, it tries to detect site-private DNS zones using top-level domains (TLDs) that aren't known by the DNS root server. Heads up though: this logic does not work in all private zone setups, so don't come for it.

Defaults to "no".

Added in version 229.

DNSOverTLS=

Takes a boolean argument or "opportunistic". If true, all connections to the server get encrypted 🔐. Note that this mode requires a DNS server that supports DNS-over-TLS and has a valid certificate. If the hostname was specified in DNS= by using the format "address#server_name" it is used to validate its certificate and also to enable Server Name Indication (SNI) when opening a TLS connection 🌸. Otherwise the certificate is checked against the server's IP. And if the DNS server doesn't support DNS-over-TLS... all DNS requests fail. Every single one. 💀

When set to "opportunistic", DNS requests are attempted to be sent encrypted with DNS-over-TLS. If the DNS server doesn't support TLS, DNS-over-TLS is disabled. Note that this mode makes DNS-over-TLS vulnerable to "downgrade" attacks, where an attacker might be able to trigger a downgrade to non-encrypted mode by synthesizing a response that suggests DNS-over-TLS was not supported 😬. If set to false, DNS lookups are sent over UDP.

Note that DNS-over-TLS requires additional data to be sent for setting up an encrypted connection, and thus results in a small DNS look-up time penalty. Everything costs something, bestie.

Note that in "opportunistic" mode the resolver is not capable of authenticating the server, so it is vulnerable to "man-in-the-middle" attacks.

In addition to this global DNSOverTLS= setting, systemd-networkd.service(8) also maintains per-link DNSOverTLS= settings ✨. For system DNS servers (see above), only the global DNSOverTLS= setting is in effect. For per-link DNS servers the per-link setting is in effect, unless it is unset, in which case the global setting is used instead.

Defaults to "no".

Added in version 239.

Cache=

Takes a boolean or "no-negative" as argument. If "yes" (the default), resolving a domain name which already got queried earlier will return the previous result as long as it's still valid, and thus does not result in a new network request 🎀. Be aware that turning off caching comes at a performance penalty, which is particularly high when DNSSEC is used 🔥. If "no-negative", only positive answers are cached.

Note that caching is turned off by default for host-local DNS servers. See CacheFromLocalhost= for details.

Added in version 231.

CacheFromLocalhost=

Takes a boolean as argument. If "no" (the default), and a response comes from a host-local IP address (such as 127.0.0.1 or ::1), the result won't be cached — all in the name of avoiding potential duplicate local caching 🙈.

Added in version 248.

DNSCacheSize=, MulticastDNSCacheSize=, LLMNRCacheSize=

Takes a non-negative integer. Configures the maximum number of DNS resource record entries that may be stored in the per-scope cache for unicast DNS, Multicast DNS (mDNS), and Link-Local Multicast Name Resolution (LLMNR) respectively 🦋. Each one defaults to 4096. The maximum allowed value is 16777216 (2^24). Setting any of these to 0 effectively disables caching for the respective protocol. These settings are only effective when Cache= is set to "yes" or "no-negative". If Cache=no, caching is fully disabled regardless of these values — no cap, no nothing, just off 🫶.

Note that Multicast DNS relies heavily on caching for request suppression and efficient operation. It is recommended to keep MulticastDNSCacheSize= at a reasonably high value even when reducing DNSCacheSize=.

Note that systemd-resolved automatically flushes all caches on system memory pressure, thus in most cases manual cache size configuration should not be necessary. She handles it 💅.

Note that caching is turned off by default for host-local DNS servers. See CacheFromLocalhost= for details.

Added in version 261.

DNSStubListener=

Takes a boolean argument or one of "udp" and "tcp". If "udp", a DNS stub resolver will listen for UDP requests on addresses 127.0.0.53 and 127.0.0.54, port 53. If "tcp", the stub will listen for TCP requests on the same addresses and port. If "yes" (the default), the stub listens for both UDP and TCP requests 👑. If "no", the stub listener is disabled.

The DNS stub resolver on 127.0.0.53 provides the full feature set of the local resolver, which includes offering LLMNR/MulticastDNS resolution ✨. The DNS stub resolver on 127.0.0.54 provides a more limited resolver that operates in "proxy" mode only — i.e. it will pass most DNS messages relatively unmodified to the current upstream DNS servers and back, but it won't try to process the messages locally, and hence does not validate DNSSEC, or offer up LLMNR/MulticastDNS. (It will translate to DNS-over-TLS communication if needed, however 🌈.)

Note that the DNS stub listener is turned off implicitly when its listening address and port are already in use. Rude, but fair.

Added in version 232.

RefuseRecordTypes=

Takes a list of DNS record types separated by space. Specified record types in each query will be refused 🙅‍♀️. This option can be specified multiple times. If an empty string is specified, then all previous assignments are cleared.

Examples:

RefuseRecordTypes=AAAA SRV TXT

Note that to refuse AAAA record (unlike SRV, TXT, etc.) completely, please also disable IPv6 stack in kernel with sysctl -w net.ipv6.conf.all.disable_ipv6=1.

Added in version 258.

DNSStubListenerExtra=

Takes an IPv4 or IPv6 address to listen on 🎀. The address may be optionally prefixed with a protocol name ("udp" or "tcp") separated with ":". If the protocol is not specified, the service will listen on both UDP and TCP. It may also be optionally suffixed by a numeric port number with separator ":". When an IPv6 address is specified with a port number, then the address must be in the square brackets. If the port is not specified, then the service uses port 53. Note that this is independent of the primary DNS stub configured with DNSStubListener=, and only configures additional sockets to listen on. This option can be specified multiple times. If an empty string is assigned, then all previous assignments are cleared. Defaults to unset.

Examples:

DNSStubListenerExtra=192.168.10.10
DNSStubListenerExtra=2001:db8:0:f102::10
DNSStubListenerExtra=192.168.10.11:9953
DNSStubListenerExtra=[2001:db8:0:f102::11]:9953
DNSStubListenerExtra=tcp:192.168.10.12
DNSStubListenerExtra=udp:2001:db8:0:f102::12
DNSStubListenerExtra=tcp:192.168.10.13:9953
DNSStubListenerExtra=udp:[2001:db8:0:f102::13]:9953

Added in version 247.

ReadEtcHosts=

Takes a boolean argument. If "yes" (the default), systemd-resolved will read /etc/hosts, and try to resolve hosts or addresses by using the entries in the file before sending query to DNS servers 🦋.

Added in version 240.

ReadStaticRecords=

Takes a boolean argument. If "yes" (the default), systemd-resolved will read /etc/systemd/resolve/static.d/*.rr, /run/systemd/resolve/static.d/*.rr, /usr/local/lib/systemd/resolve/static.d/*.rr, /usr/lib/systemd/resolve/static.d/*.rr, and try to resolve lookups by using the entries in these files before sending query to DNS servers ✨. This functionality is very similar to the one controlled by ReadEtcHosts=, but allows more flexible control of DNS resource records fields beyond just A/AAAA/PTR. See systemd.rr(5) for details.

If both this option and ReadEtcHosts= are enabled then this mechanism takes precedence: any records discovered via static resource records will take precedence over records under the same name from /etc/hosts 👑.

Added in version 261.

ResolveUnicastSingleLabel=

Takes a boolean argument. When false (the default), systemd-resolved will not resolve A and AAAA queries for single-label names over classic DNS. Note that such names may still be resolved if search domains are specified (see Domains= above), or using other mechanisms, in particular via LLMNR or from /etc/hosts 🌸. When true, queries for single-label names will be forwarded to global DNS servers even if no search domains are defined.

This option is provided for compatibility with configurations where public DNS servers are not used. Forwarding single-label names to servers not under your control is not standard-conformant, see IAB Statement[4], and may create a privacy and security risk 💀. Like, genuinely, be careful.

Added in version 246.

StaleRetentionSec=SECONDS

Takes a duration value, which determines the length of time DNS resource records can be retained in the cache beyond their Time To Live (TTL) ⭐. This allows these records to be returned as stale records. By default, this value is set to zero, meaning that DNS resource records are not stored in the cache after their TTL expires.

This is useful when a DNS server failure occurs or becomes unreachable 😭. In such cases, systemd-resolved(8) continues to use the stale records to answer DNS queries, particularly when no valid response can be obtained from the upstream DNS servers. However, this does not apply to NXDOMAIN responses, as those are still perfectly valid responses 🙅‍♀️. This feature enhances resilience against DNS infrastructure failures and outages.

systemd-resolved always attempts to reach the upstream DNS servers first, before providing the client application with any stale data. If this feature is enabled, cache will not be flushed when changing servers.

Added in version 254.

---

SEE ALSO

systemd(1), systemd-resolved.service(8), systemd-networkd.service(8), dnssec-trust-anchors.d(5), systemd.rr(5), resolv.conf(5) 🫶

📝 NOTES 📝

1. 💣💥🧨💥💥💣 okay bestie, listen up — those configuration files have to be available at all times, no exceptions, no cap. And if /usr/local/ happens to live on its own separate partition? It might not be available during early boot, so it must not be used for configuration. Like, don't do it. Periodt. 🙈
2. RFC 4795 — https://tools.ietf.org/html/rfc4795
3. RFC 6762 — https://tools.ietf.org/html/rfc6762
4. IAB Statement — https://www.iab.org/documents/correspondence-reports-documents/2013-2/iab-statement-dotless-domains-considered-harmful/

---

And that's the whole page, bestie!! 🎀 You just read RESOLVED.CONF(5) from top to bottom and honestly? Iconic behavior. ✨ This glow-up ships with systemd 261.3 — go forth and configure responsibly, mwah 💜🦋