1unbound.conf(5) unbound 1.8.3 unbound.conf(5)
2
6 unbound.conf - Unbound configuration file.
7
9 unbound.conf
10
12 unbound.conf is used to configure unbound(8). The file format has
13 attributes and values. Some attributes have attributes inside them.
14 The notation is: attribute: value.
15 Comments start with # and last to the end of line. Empty lines are
17 ignored as is whitespace at the beginning of a line.
18 The utility unbound-checkconf(8) can be used to check unbound.conf
20 prior to usage.
21
23 An example config file is shown below. Copy this to
24 /etc/unbound/unbound.conf and start the server with:
25 $ unbound -c /etc/unbound/unbound.conf
27 Most settings are the defaults. Stop the server with:
29 $ kill `cat /etc/unbound/unbound.pid`
31 Below is a minimal config file. The source distribution contains an
33 extensive example.conf file with all the options.
34 # unbound.conf(5) config file for unbound(8).
36 server:
37 directory: "/etc/unbound"
38 username: unbound
39 # make sure unbound can access entropy from inside the chroot.
40 # e.g. on linux the use these commands (on BSD, devfs(8) is used):
41 # mount --bind -n /dev/random /etc/unbound/dev/random
42 # and mount --bind -n /dev/log /etc/unbound/dev/log
43 chroot: "/etc/unbound"
44 # logfile: "/etc/unbound/unbound.log" #uncomment to use logfile.
45 pidfile: "/etc/unbound/unbound.pid"
46 # verbosity: 1 # uncomment and increase to get more logging.
47 # listen on all interfaces, answer queries from the local subnet.
48 interface: 0.0.0.0
49 interface: ::0
50 access-control: 10.0.0.0/8 allow
51 access-control: 2001:DB8::/64 allow
52
54 There must be whitespace between keywords. Attribute keywords end with
55 a colon ':'. An attribute is followed by its containing attributes, or
56 a value.
57 Files can be included using the include: directive. It can appear any‐
59 where, it accepts a single file name as argument. Processing continues
60 as if the text from the included file was copied into the config file
61 at that point. If also using chroot, using full path names for the
62 included files works, relative pathnames for the included names work if
63 the directory where the daemon is started equals its chroot/working
64 directory or is specified before the include statement with directory:
65 dir. Wildcards can be used to include multiple files, see glob(7).
66 Server Options
68 These options are part of the server: clause.
69 verbosity: <number>
71 The verbosity number, level 0 means no verbosity, only errors.
72 Level 1 gives operational information. Level 2 gives detailed
73 operational information. Level 3 gives query level information,
74 output per query. Level 4 gives algorithm level information.
75 Level 5 logs client identification for cache misses. Default is
76 level 1. The verbosity can also be increased from the command‐
77 line, see unbound(8).
78 statistics-interval: <seconds>
80 The number of seconds between printing statistics to the log for
81 every thread. Disable with value 0 or "". Default is disabled.
82 The histogram statistics are only printed if replies were sent
83 during the statistics interval, requestlist statistics are
84 printed for every interval (but can be 0). This is because the
85 median calculation requires data to be present.
86 statistics-cumulative: <yes or no>
88 If enabled, statistics are cumulative since starting unbound,
89 without clearing the statistics counters after logging the sta‐
90 tistics. Default is no.
91 extended-statistics: <yes or no>
93 If enabled, extended statistics are printed from unbound-con‐
94 trol(8). Default is off, because keeping track of more statis‐
95 tics takes time. The counters are listed in unbound-control(8).
96 num-threads: <number>
98 The number of threads to create to serve clients. Use 1 for no
99 threading.
100 port: <port number>
102 The port number, default 53, on which the server responds to
103 queries.
104 interface: <ip address[@port]>
106 Interface to use to connect to the network. This interface is
107 listened to for queries from clients, and answers to clients are
108 given from it. Can be given multiple times to work on several
109 interfaces. If none are given the default is to listen to local‐
110 host. The interfaces are not changed on a reload (kill -HUP)
111 but only on restart. A port number can be specified with @port
112 (without spaces between interface and port number), if not spec‐
113 ified the default port (from port) is used.
114 ip-address: <ip address[@port]>
116 Same as interface: (for ease of compatibility with nsd.conf).
117 interface-automatic: <yes or no>
119 Detect source interface on UDP queries and copy them to replies.
120 This feature is experimental, and needs support in your OS for
121 particular socket options. Default value is no.
122 outgoing-interface: <ip address or ip6 netblock>
124 Interface to use to connect to the network. This interface is
125 used to send queries to authoritative servers and receive their
126 replies. Can be given multiple times to work on several inter‐
127 faces. If none are given the default (all) is used. You can
128 specify the same interfaces in interface: and outgoing-inter‐
129 face: lines, the interfaces are then used for both purposes.
130 Outgoing queries are sent via a random outgoing interface to
131 counter spoofing.
132 If an IPv6 netblock is specified instead of an individual IPv6
134 address, outgoing UDP queries will use a randomised source
135 address taken from the netblock to counter spoofing. Requires
136 the IPv6 netblock to be routed to the host running unbound, and
137 requires OS support for unprivileged non-local binds (currently
138 only supported on Linux). Several netblocks may be specified
139 with multiple outgoing-interface: options, but do not specify
140 both an individual IPv6 address and an IPv6 netblock, or the
141 randomisation will be compromised. Consider combining with pre‐
142 fer-ip6: yes to increase the likelihood of IPv6 nameservers
143 being selected for queries. On Linux you need these two com‐
144 mands to be able to use the freebind socket option to receive
145 traffic for the ip6 netblock: ip -6 addr add mynetblock/64 dev
146 lo && ip -6 route add local mynetblock/64 dev lo
147 outgoing-range: <number>
149 Number of ports to open. This number of file descriptors can be
150 opened per thread. Must be at least 1. Default depends on com‐
151 pile options. Larger numbers need extra resources from the oper‐
152 ating system. For performance a very large value is best, use
153 libevent to make this possible.
154 outgoing-port-permit: <port number or range>
156 Permit unbound to open this port or range of ports for use to
157 send queries. A larger number of permitted outgoing ports
158 increases resilience against spoofing attempts. Make sure these
159 ports are not needed by other daemons. By default only ports
160 above 1024 that have not been assigned by IANA are used. Give a
161 port number or a range of the form "low-high", without spaces.
162 The outgoing-port-permit and outgoing-port-avoid statements are
164 processed in the line order of the config file, adding the per‐
165 mitted ports and subtracting the avoided ports from the set of
166 allowed ports. The processing starts with the non IANA allo‐
167 cated ports above 1024 in the set of allowed ports.
168 outgoing-port-avoid: <port number or range>
170 Do not permit unbound to open this port or range of ports for
171 use to send queries. Use this to make sure unbound does not grab
172 a port that another daemon needs. The port is avoided on all
173 outgoing interfaces, both IP4 and IP6. By default only ports
174 above 1024 that have not been assigned by IANA are used. Give a
175 port number or a range of the form "low-high", without spaces.
176 outgoing-num-tcp: <number>
178 Number of outgoing TCP buffers to allocate per thread. Default
179 is 10. If set to 0, or if do-tcp is "no", no TCP queries to
180 authoritative servers are done. For larger installations
181 increasing this value is a good idea.
182 incoming-num-tcp: <number>
184 Number of incoming TCP buffers to allocate per thread. Default
185 is 10. If set to 0, or if do-tcp is "no", no TCP queries from
186 clients are accepted. For larger installations increasing this
187 value is a good idea.
188 edns-buffer-size: <number>
190 Number of bytes size to advertise as the EDNS reassembly buffer
191 size. This is the value put into datagrams over UDP towards
192 peers. The actual buffer size is determined by msg-buffer-size
193 (both for TCP and UDP). Do not set higher than that value.
194 Default is 4096 which is RFC recommended. If you have fragmen‐
195 tation reassembly problems, usually seen as timeouts, then a
196 value of 1472 can fix it. Setting to 512 bypasses even the most
197 stringent path MTU problems, but is seen as extreme, since the
198 amount of TCP fallback generated is excessive (probably also for
199 this resolver, consider tuning the outgoing tcp number).
200 max-udp-size: <number>
202 Maximum UDP response size (not applied to TCP response). 65536
203 disables the udp response size maximum, and uses the choice from
204 the client, always. Suggested values are 512 to 4096. Default
205 is 4096.
206 msg-buffer-size: <number>
208 Number of bytes size of the message buffers. Default is 65552
209 bytes, enough for 64 Kb packets, the maximum DNS message size.
210 No message larger than this can be sent or received. Can be
211 reduced to use less memory, but some requests for DNS data, such
212 as for huge resource records, will result in a SERVFAIL reply to
213 the client.
214 msg-cache-size: <number>
216 Number of bytes size of the message cache. Default is 4
217 megabytes. A plain number is in bytes, append 'k', 'm' or 'g'
218 for kilobytes, megabytes or gigabytes (1024*1024 bytes in a
219 megabyte).
220 msg-cache-slabs: <number>
222 Number of slabs in the message cache. Slabs reduce lock con‐
223 tention by threads. Must be set to a power of 2. Setting
224 (close) to the number of cpus is a reasonable guess.
225 num-queries-per-thread: <number>
227 The number of queries that every thread will service simultane‐
228 ously. If more queries arrive that need servicing, and no
229 queries can be jostled out (see jostle-timeout), then the
230 queries are dropped. This forces the client to resend after a
231 timeout; allowing the server time to work on the existing
232 queries. Default depends on compile options, 512 or 1024.
233 jostle-timeout: <msec>
235 Timeout used when the server is very busy. Set to a value that
236 usually results in one roundtrip to the authority servers. If
237 too many queries arrive, then 50% of the queries are allowed to
238 run to completion, and the other 50% are replaced with the new
239 incoming query if they have already spent more than their
240 allowed time. This protects against denial of service by slow
241 queries or high query rates. Default 200 milliseconds. The
242 effect is that the qps for long-lasting queries is about (num‐
243 queriesperthread / 2) / (average time for such long queries)
244 qps. The qps for short queries can be about (numqueries‐
245 perthread / 2) / (jostletimeout in whole seconds) qps per
246 thread, about (1024/2)*5 = 2560 qps by default.
247 delay-close: <msec>
249 Extra delay for timeouted UDP ports before they are closed, in
250 msec. Default is 0, and that disables it. This prevents very
251 delayed answer packets from the upstream (recursive) servers
252 from bouncing against closed ports and setting off all sort of
253 close-port counters, with eg. 1500 msec. When timeouts happen
254 you need extra sockets, it checks the ID and remote IP of pack‐
255 ets, and unwanted packets are added to the unwanted packet
256 counter.
257 unknown-server-time-limit: <msec>
259 The wait time in msec for waiting for an unknown server to
260 reply. Increase this if you are behind a slow satellite link,
261 to eg. 1128. That would then avoid re-querying every initial
262 query because it times out. Default is 376 msec.
263 so-rcvbuf: <number>
265 If not 0, then set the SO_RCVBUF socket option to get more buf‐
266 fer space on UDP port 53 incoming queries. So that short spikes
267 on busy servers do not drop packets (see counter in netstat
268 -su). Default is 0 (use system value). Otherwise, the number
269 of bytes to ask for, try "4m" on a busy server. The OS caps it
270 at a maximum, on linux unbound needs root permission to bypass
271 the limit, or the admin can use sysctl net.core.rmem_max. On
272 BSD change kern.ipc.maxsockbuf in /etc/sysctl.conf. On OpenBSD
273 change header and recompile kernel. On Solaris ndd -set /dev/udp
274 udp_max_buf 8388608.
275 so-sndbuf: <number>
277 If not 0, then set the SO_SNDBUF socket option to get more buf‐
278 fer space on UDP port 53 outgoing queries. This for very busy
279 servers handles spikes in answer traffic, otherwise 'send:
280 resource temporarily unavailable' can get logged, the buffer
281 overrun is also visible by netstat -su. Default is 0 (use sys‐
282 tem value). Specify the number of bytes to ask for, try "4m" on
283 a very busy server. The OS caps it at a maximum, on linux
284 unbound needs root permission to bypass the limit, or the admin
285 can use sysctl net.core.wmem_max. On BSD, Solaris changes are
286 similar to so-rcvbuf.
287 so-reuseport: <yes or no>
289 If yes, then open dedicated listening sockets for incoming
290 queries for each thread and try to set the SO_REUSEPORT socket
291 option on each socket. May distribute incoming queries to
292 threads more evenly. Default is yes. On Linux it is supported
293 in kernels >= 3.9. On other systems, FreeBSD, OSX it may also
294 work. You can enable it (on any platform and kernel), it then
295 attempts to open the port and passes the option if it was avail‐
296 able at compile time, if that works it is used, if it fails, it
297 continues silently (unless verbosity 3) without the option.
298 ip-transparent: <yes or no>
300 If yes, then use IP_TRANSPARENT socket option on sockets where
301 unbound is listening for incoming traffic. Default no. Allows
302 you to bind to non-local interfaces. For example for non-exis‐
303 tent IP addresses that are going to exist later on, with host
304 failover configuration. This is a lot like interface-automatic,
305 but that one services all interfaces and with this option you
306 can select which (future) interfaces unbound provides service
307 on. This option needs unbound to be started with root permis‐
308 sions on some systems. The option uses IP_BINDANY on FreeBSD
309 systems and SO_BINDANY on OpenBSD systems.
310 ip-freebind: <yes or no>
312 If yes, then use IP_FREEBIND socket option on sockets where
313 unbound is listening to incoming traffic. Default no. Allows
314 you to bind to IP addresses that are nonlocal or do not exist,
315 like when the network interface or IP address is down. Exists
316 only on Linux, where the similar ip-transparent option is also
317 available.
318 rrset-cache-size: <number>
320 Number of bytes size of the RRset cache. Default is 4 megabytes.
321 A plain number is in bytes, append 'k', 'm' or 'g' for kilo‐
322 bytes, megabytes or gigabytes (1024*1024 bytes in a megabyte).
323 rrset-cache-slabs: <number>
325 Number of slabs in the RRset cache. Slabs reduce lock contention
326 by threads. Must be set to a power of 2.
327 cache-max-ttl: <seconds>
329 Time to live maximum for RRsets and messages in the cache.
330 Default is 86400 seconds (1 day). If the maximum kicks in,
331 responses to clients still get decrementing TTLs based on the
332 original (larger) values. When the internal TTL expires, the
333 cache item has expired. Can be set lower to force the resolver
334 to query for data often, and not trust (very large) TTL values.
335 cache-min-ttl: <seconds>
337 Time to live minimum for RRsets and messages in the cache.
338 Default is 0. If the minimum kicks in, the data is cached for
339 longer than the domain owner intended, and thus less queries are
340 made to look up the data. Zero makes sure the data in the cache
341 is as the domain owner intended, higher values, especially more
342 than an hour or so, can lead to trouble as the data in the cache
343 does not match up with the actual data any more.
344 cache-max-negative-ttl: <seconds>
346 Time to live maximum for negative responses, these have a SOA in
347 the authority section that is limited in time. Default is 3600.
348 This applies to nxdomain and nodata answers.
349 infra-host-ttl: <seconds>
351 Time to live for entries in the host cache. The host cache con‐
352 tains roundtrip timing, lameness and EDNS support information.
353 Default is 900.
354 infra-cache-slabs: <number>
356 Number of slabs in the infrastructure cache. Slabs reduce lock
357 contention by threads. Must be set to a power of 2.
358 infra-cache-numhosts: <number>
360 Number of hosts for which information is cached. Default is
361 10000.
362 infra-cache-min-rtt: <msec>
364 Lower limit for dynamic retransmit timeout calculation in infra‐
365 structure cache. Default is 50 milliseconds. Increase this value
366 if using forwarders needing more time to do recursive name reso‐
367 lution.
368 define-tag: <"list of tags">
370 Define the tags that can be used with local-zone and access-con‐
371 trol. Enclose the list between quotes ("") and put spaces
372 between tags.
373 do-ip4: <yes or no>
375 Enable or disable whether ip4 queries are answered or issued.
376 Default is yes.
377 do-ip6: <yes or no>
379 Enable or disable whether ip6 queries are answered or issued.
380 Default is yes. If disabled, queries are not answered on IPv6,
381 and queries are not sent on IPv6 to the internet nameservers.
382 With this option you can disable the ipv6 transport for sending
383 DNS traffic, it does not impact the contents of the DNS traffic,
384 which may have ip4 and ip6 addresses in it.
385 prefer-ip6: <yes or no>
387 If enabled, prefer IPv6 transport for sending DNS queries to
388 internet nameservers. Default is no.
389 do-udp: <yes or no>
391 Enable or disable whether UDP queries are answered or issued.
392 Default is yes.
393 do-tcp: <yes or no>
395 Enable or disable whether TCP queries are answered or issued.
396 Default is yes.
397 tcp-mss: <number>
399 Maximum segment size (MSS) of TCP socket on which the server
400 responds to queries. Value lower than common MSS on Ethernet
401 (1220 for example) will address path MTU problem. Note that not
402 all platform supports socket option to set MSS (TCP_MAXSEG).
403 Default is system default MSS determined by interface MTU and
404 negotiation between server and client.
405 outgoing-tcp-mss: <number>
407 Maximum segment size (MSS) of TCP socket for outgoing queries
408 (from Unbound to other servers). Value lower than common MSS on
409 Ethernet (1220 for example) will address path MTU problem. Note
410 that not all platform supports socket option to set MSS
411 (TCP_MAXSEG). Default is system default MSS determined by
412 interface MTU and negotiation between Unbound and other servers.
413 tcp-idle-timeout: <msec>
415 The period Unbound will wait for a query on a TCP connection.
416 If this timeout expires Unbound closes the connection. This
417 option defaults to 30000 milliseconds. When the number of free
418 incoming TCP buffers falls below 50% of the total number config‐
419 ured, the option value used is progressively reduced, first to
420 1% of the configured value, then to 0.2% of the configured value
421 if the number of free buffers falls below 35% of the total num‐
422 ber configured, and finally to 0 if the number of free buffers
423 falls below 20% of the total number configured. A minimum time‐
424 out of 200 milliseconds is observed regardless of the option
425 value used.
426 edns-tcp-keepalive: <yes or no>
428 Enable or disable EDNS TCP Keepalive. Default is no.
429 edns-tcp-keepalive-timeout: <msec>
431 The period Unbound will wait for a query on a TCP connection
432 when EDNS TCP Keepalive is active. If this timeout expires
433 Unbound closes the connection. If the client supports the EDNS
434 TCP Keepalive option, Unbound sends the timeout value to the
435 client to encourage it to close the connection before the server
436 times out. This option defaults to 120000 milliseconds. When
437 the number of free incoming TCP buffers falls below 50% of the
438 total number configured, the advertised timeout is progressively
439 reduced to 1% of the configured value, then to 0.2% of the con‐
440 figured value if the number of free buffers falls below 35% of
441 the total number configured, and finally to 0 if the number of
442 free buffers falls below 20% of the total number configured. A
443 minimum actual timeout of 200 milliseconds is observed regard‐
444 less of the advertised timeout.
445 tcp-upstream: <yes or no>
447 Enable or disable whether the upstream queries use TCP only for
448 transport. Default is no. Useful in tunneling scenarios.
449 udp-upstream-without-downstream: <yes or no>
451 Enable udp upstream even if do-udp is no. Default is no, and
452 this does not change anything. Useful for TLS service
453 providers, that want no udp downstream but use udp to fetch data
454 upstream.
455 tls-upstream: <yes or no>
457 Enabled or disable whether the upstream queries use TLS only for
458 transport. Default is no. Useful in tunneling scenarios. The
459 TLS contains plain DNS in TCP wireformat. The other server must
460 support this (see tls-service-key). If you enable this, also
461 configure a tls-cert-bundle or use tls-win-cert to load CA
462 certs, otherwise the connections cannot be authenticated.
463 ssl-upstream: <yes or no>
465 Alternate syntax for tls-upstream. If both are present in the
466 config file the last is used.
467 tls-service-key: <file>
469 If enabled, the server provider TLS service on its TCP sockets.
470 The clients have to use tls-upstream: yes. The file is the pri‐
471 vate key for the TLS session. The public certificate is in the
472 tls-service-pem file. Default is "", turned off. Requires a
473 restart (a reload is not enough) if changed, because the private
474 key is read while root permissions are held and before chroot
475 (if any). Normal DNS TCP service is not provided and gives
476 errors, this service is best run with a different port: config
477 or @port suffixes in the interface config.
478 ssl-service-key: <file>
480 Alternate syntax for tls-service-key.
481 tls-service-pem: <file>
483 The public key certificate pem file for the tls service.
484 Default is "", turned off.
485 ssl-service-pem: <file>
487 Alternate syntax for tls-service-pem.
488 tls-port: <number>
490 The port number on which to provide TCP TLS service, default
491 853, only interfaces configured with that port number as @number
492 get the TLS service.
493 ssl-port: <number>
495 Alternate syntax for tls-port.
496 tls-cert-bundle: <file>
498 If null or "", no file is used. Set it to the certificate bun‐
499 dle file, for example "/etc/pki/tls/certs/ca-bundle.crt". These
500 certificates are used for authenticating connections made to
501 outside peers. For example auth-zone urls, and also DNS over
502 TLS connections.
503 ssl-cert-bundle: <file>
505 Alternate syntax for tls-cert-bundle.
506 tls-win-cert: <yes or no>
508 Add the system certificates to the cert bundle certificates for
509 authentication. If no cert bundle, it uses only these certifi‐
510 cates. Default is no. On windows this option uses the certifi‐
511 cates from the cert store. Use the tls-cert-bundle option on
512 other systems.
513 tls-additional-port: <portnr>
515 List portnumbers as tls-additional-port, and when interfaces are
516 defined, eg. with the @port suffix, as this port number, they
517 provide dns over TLS service. Can list multiple, each on a new
518 statement.
519 use-systemd: <yes or no>
521 Enable or disable systemd socket activation. Default is no.
522 do-daemonize: <yes or no>
524 Enable or disable whether the unbound server forks into the
525 background as a daemon. Set the value to no when unbound runs
526 as systemd service. Default is yes.
527 tcp-connection-limit: <IP netblock> <limit>
529 Allow up to limit simultaneous TCP connections from the given
530 netblock. When at the limit, further connections are accepted
531 but closed immediately. This option is experimental at this
532 time.
533 access-control: <IP netblock> <action>
535 The netblock is given as an IP4 or IP6 address with /size
536 appended for a classless network block. The action can be deny,
537 refuse, allow, allow_setrd, allow_snoop, deny_non_local or
538 refuse_non_local. The most specific netblock match is used, if
539 none match deny is used.
540 The action deny stops queries from hosts from that netblock.
542 The action refuse stops queries too, but sends a DNS rcode
544 REFUSED error message back.
545 The action allow gives access to clients from that netblock. It
547 gives only access for recursion clients (which is what almost
548 all clients need). Nonrecursive queries are refused.
549 The allow action does allow nonrecursive queries to access the
551 local-data that is configured. The reason is that this does not
552 involve the unbound server recursive lookup algorithm, and
553 static data is served in the reply. This supports normal opera‐
554 tions where nonrecursive queries are made for the authoritative
555 data. For nonrecursive queries any replies from the dynamic
556 cache are refused.
557 The allow_setrd action ignores the recursion desired (RD) bit
559 and treats all requests as if the recursion desired bit is set.
560 Note that this behavior violates RFC 1034 which states that a
561 name server should never perform recursive service unless asked
562 via the RD bit since this interferes with trouble shooting of
563 name servers and their databases. This prohibited behavior may
564 be useful if another DNS server must forward requests for spe‐
565 cific zones to a resolver DNS server, but only supports stub
566 domains and sends queries to the resolver DNS server with the RD
567 bit cleared.
568 The action allow_snoop gives nonrecursive access too. This give
570 both recursive and non recursive access. The name allow_snoop
571 refers to cache snooping, a technique to use nonrecursive
572 queries to examine the cache contents (for malicious acts).
573 However, nonrecursive queries can also be a valuable debugging
574 tool (when you want to examine the cache contents). In that case
575 use allow_snoop for your administration host.
576 By default only localhost is allowed, the rest is refused. The
578 default is refused, because that is protocol-friendly. The DNS
579 protocol is not designed to handle dropped packets due to pol‐
580 icy, and dropping may result in (possibly excessive) retried
581 queries.
582 The deny_non_local and refuse_non_local settings are for hosts
584 that are only allowed to query for the authoritative local-data,
585 they are not allowed full recursion but only the static data.
586 With deny_non_local, messages that are disallowed are dropped,
587 with refuse_non_local they receive error code REFUSED.
588 access-control-tag: <IP netblock> <"list of tags">
590 Assign tags to access-control elements. Clients using this
591 access control element use localzones that are tagged with one
592 of these tags. Tags must be defined in define-tags. Enclose
593 list of tags in quotes ("") and put spaces between tags. If
594 access-control-tag is configured for a netblock that does not
595 have an access-control, an access-control element with action
596 allow is configured for this netblock.
597 access-control-tag-action: <IP netblock> <tag> <action>
599 Set action for particular tag for given access control element.
600 If you have multiple tag values, the tag used to lookup the
601 action is the first tag match between access-control-tag and
602 local-zone-tag where "first" comes from the order of the define-
603 tag values.
604 access-control-tag-data: <IP netblock> <tag> <"resource record string">
606 Set redirect data for particular tag for given access control
607 element.
608 access-control-view: <IP netblock> <view name>
610 Set view for given access control element.
611 chroot: <directory>
613 If chroot is enabled, you should pass the configfile (from the
614 commandline) as a full path from the original root. After the
615 chroot has been performed the now defunct portion of the config
616 file path is removed to be able to reread the config after a
617 reload.
618 All other file paths (working dir, logfile, roothints, and key
620 files) can be specified in several ways: as an absolute path
621 relative to the new root, as a relative path to the working
622 directory, or as an absolute path relative to the original root.
623 In the last case the path is adjusted to remove the unused por‐
624 tion.
625 The pidfile can be either a relative path to the working direc‐
627 tory, or an absolute path relative to the original root. It is
628 written just prior to chroot and dropping permissions. This
629 allows the pidfile to be /var/run/unbound.pid and the chroot to
630 be /var/unbound, for example.
631 Additionally, unbound may need to access /dev/random (for
633 entropy) from inside the chroot.
634 If given a chroot is done to the given directory. By default
636 chroot is enabled and the default is "/etc/unbound". If you give
637 "" no chroot is performed.
638 username: <name>
640 If given, after binding the port the user privileges are
641 dropped. Default is "unbound". If you give username: "" no user
642 change is performed.
643 If this user is not capable of binding the port, reloads (by
645 signal HUP) will still retain the opened ports. If you change
646 the port number in the config file, and that new port number
647 requires privileges, then a reload will fail; a restart is
648 needed.
649 directory: <directory>
651 Sets the working directory for the program. Default is
652 "/etc/unbound". On Windows the string "%EXECUTABLE%" tries to
653 change to the directory that unbound.exe resides in. If you
654 give a server: directory: dir before include: file statements
655 then those includes can be relative to the working directory.
656 logfile: <filename>
658 If "" is given, logging goes to stderr, or nowhere once daemo‐
659 nized. The logfile is appended to, in the following format:
660 [seconds since 1970] unbound[pid:tid]: type: message.
661 If this option is given, the use-syslog is option is set to
662 "no". The logfile is reopened (for append) when the config file
663 is reread, on SIGHUP.
664 use-syslog: <yes or no>
666 Sets unbound to send log messages to the syslogd, using sys‐
667 log(3). The log facility LOG_DAEMON is used, with identity
668 "unbound". The logfile setting is overridden when use-syslog is
669 turned on. The default is to log to syslog.
670 log-identity: <string>
672 If "" is given (default), then the name of the executable, usu‐
673 ally "unbound" is used to report to the log. Enter a string to
674 override it with that, which is useful on systems that run more
675 than one instance of unbound, with different configurations, so
676 that the logs can be easily distinguished against.
677 log-time-ascii: <yes or no>
679 Sets logfile lines to use a timestamp in UTC ascii. Default is
680 no, which prints the seconds since 1970 in brackets. No effect
681 if using syslog, in that case syslog formats the timestamp
682 printed into the log files.
683 log-queries: <yes or no>
685 Prints one line per query to the log, with the log timestamp and
686 IP address, name, type and class. Default is no. Note that it
687 takes time to print these lines which makes the server (signifi‐
688 cantly) slower. Odd (nonprintable) characters in names are
689 printed as '?'.
690 log-replies: <yes or no>
692 Prints one line per reply to the log, with the log timestamp and
693 IP address, name, type, class, return code, time to resolve,
694 from cache and response size. Default is no. Note that it
695 takes time to print these lines which makes the server (signifi‐
696 cantly) slower. Odd (nonprintable) characters in names are
697 printed as '?'.
698 log-local-actions: <yes or no>
700 Print log lines to inform about local zone actions. These lines
701 are like the local-zone type inform prints out, but they are
702 also printed for the other types of local zones.
703 log-servfail: <yes or no>
705 Print log lines that say why queries return SERVFAIL to clients.
706 This is separate from the verbosity debug logs, much smaller,
707 and printed at the error level, not the info level of debug info
708 from verbosity.
709 pidfile: <filename>
711 The process id is written to the file. Default is
712 "/var/run/unbound/unbound.pid". So,
713 kill -HUP `cat /var/run/unbound/unbound.pid`
714 triggers a reload,
715 kill -TERM `cat /var/run/unbound/unbound.pid`
716 gracefully terminates.
717 root-hints: <filename>
719 Read the root hints from this file. Default is nothing, using
720 builtin hints for the IN class. The file has the format of zone
721 files, with root nameserver names and addresses only. The
722 default may become outdated, when servers change, therefore it
723 is good practice to use a root-hints file.
724 hide-identity: <yes or no>
726 If enabled id.server and hostname.bind queries are refused.
727 identity: <string>
729 Set the identity to report. If set to "", the default, then the
730 hostname of the server is returned.
731 hide-version: <yes or no>
733 If enabled version.server and version.bind queries are refused.
734 version: <string>
736 Set the version to report. If set to "", the default, then the
737 package version is returned.
738 hide-trustanchor: <yes or no>
740 If enabled trustanchor.unbound queries are refused.
741 target-fetch-policy: <"list of numbers">
743 Set the target fetch policy used by unbound to determine if it
744 should fetch nameserver target addresses opportunistically. The
745 policy is described per dependency depth.
746 The number of values determines the maximum dependency depth
748 that unbound will pursue in answering a query. A value of -1
749 means to fetch all targets opportunistically for that dependency
750 depth. A value of 0 means to fetch on demand only. A positive
751 value fetches that many targets opportunistically.
752 Enclose the list between quotes ("") and put spaces between num‐
754 bers. The default is "3 2 1 0 0". Setting all zeroes, "0 0 0 0
755 0" gives behaviour closer to that of BIND 9, while setting "-1
756 -1 -1 -1 -1" gives behaviour rumoured to be closer to that of
757 BIND 8.
758 harden-short-bufsize: <yes or no>
760 Very small EDNS buffer sizes from queries are ignored. Default
761 is off, since it is legal protocol wise to send these, and
762 unbound tries to give very small answers to these queries, where
763 possible.
764 harden-large-queries: <yes or no>
766 Very large queries are ignored. Default is off, since it is
767 legal protocol wise to send these, and could be necessary for
768 operation if TSIG or EDNS payload is very large.
769 harden-glue: <yes or no>
771 Will trust glue only if it is within the servers authority.
772 Default is on.
773 harden-dnssec-stripped: <yes or no>
775 Require DNSSEC data for trust-anchored zones, if such data is
776 absent, the zone becomes bogus. If turned off, and no DNSSEC
777 data is received (or the DNSKEY data fails to validate), then
778 the zone is made insecure, this behaves like there is no trust
779 anchor. You could turn this off if you are sometimes behind an
780 intrusive firewall (of some sort) that removes DNSSEC data from
781 packets, or a zone changes from signed to unsigned to badly
782 signed often. If turned off you run the risk of a downgrade
783 attack that disables security for a zone. Default is on.
784 harden-below-nxdomain: <yes or no>
786 From RFC 8020 (with title "NXDOMAIN: There Really Is Nothing
787 Underneath"), returns nxdomain to queries for a name below
788 another name that is already known to be nxdomain. DNSSEC man‐
789 dates noerror for empty nonterminals, hence this is possible.
790 Very old software might return nxdomain for empty nonterminals
791 (that usually happen for reverse IP address lookups), and thus
792 may be incompatible with this. To try to avoid this only
793 DNSSEC-secure nxdomains are used, because the old software does
794 not have DNSSEC. Default is on. The nxdomain must be secure,
795 this means nsec3 with optout is insufficient.
796 harden-referral-path: <yes or no>
798 Harden the referral path by performing additional queries for
799 infrastructure data. Validates the replies if trust anchors are
800 configured and the zones are signed. This enforces DNSSEC vali‐
801 dation on nameserver NS sets and the nameserver addresses that
802 are encountered on the referral path to the answer. Default no,
803 because it burdens the authority servers, and it is not RFC
804 standard, and could lead to performance problems because of the
805 extra query load that is generated. Experimental option. If
806 you enable it consider adding more numbers after the tar‐
807 get-fetch-policy to increase the max depth that is checked to.
808 harden-algo-downgrade: <yes or no>
810 Harden against algorithm downgrade when multiple algorithms are
811 advertised in the DS record. If no, allows the weakest algo‐
812 rithm to validate the zone. Default is no. Zone signers must
813 produce zones that allow this feature to work, but sometimes
814 they do not, and turning this option off avoids that validation
815 failure.
816 use-caps-for-id: <yes or no>
818 Use 0x20-encoded random bits in the query to foil spoof
819 attempts. This perturbs the lowercase and uppercase of query
820 names sent to authority servers and checks if the reply still
821 has the correct casing. Disabled by default. This feature is
822 an experimental implementation of draft dns-0x20.
823 caps-whitelist: <domain>
825 Whitelist the domain so that it does not receive caps-for-id
826 perturbed queries. For domains that do not support 0x20 and
827 also fail with fallback because they keep sending different
828 answers, like some load balancers. Can be given multiple times,
829 for different domains.
830 qname-minimisation: <yes or no>
832 Send minimum amount of information to upstream servers to
833 enhance privacy. Only sent minimum required labels of the QNAME
834 and set QTYPE to A when possible. Best effort approach; full
835 QNAME and original QTYPE will be sent when upstream replies with
836 a RCODE other than NOERROR, except when receiving NXDOMAIN from
837 a DNSSEC signed zone. Default is yes.
838 qname-minimisation-strict: <yes or no>
840 QNAME minimisation in strict mode. Do not fall-back to sending
841 full QNAME to potentially broken nameservers. A lot of domains
842 will not be resolvable when this option in enabled. Only use if
843 you know what you are doing. This option only has effect when
844 qname-minimisation is enabled. Default is off.
845 aggressive-nsec: <yes or no>
847 Aggressive NSEC uses the DNSSEC NSEC chain to synthesize NXDO‐
848 MAIN and other denials, using information from previous NXDO‐
849 MAINs answers. Default is no. It helps to reduce the query
850 rate towards targets that get a very high nonexistent name
851 lookup rate.
852 private-address: <IP address or subnet>
854 Give IPv4 of IPv6 addresses or classless subnets. These are
855 addresses on your private network, and are not allowed to be
856 returned for public internet names. Any occurrence of such
857 addresses are removed from DNS answers. Additionally, the DNSSEC
858 validator may mark the answers bogus. This protects against
859 so-called DNS Rebinding, where a user browser is turned into a
860 network proxy, allowing remote access through the browser to
861 other parts of your private network. Some names can be allowed
862 to contain your private addresses, by default all the local-data
863 that you configured is allowed to, and you can specify addi‐
864 tional names using private-domain. No private addresses are
865 enabled by default. We consider to enable this for the RFC1918
866 private IP address space by default in later releases. That
867 would enable private addresses for 10.0.0.0/8 172.16.0.0/12
868 192.168.0.0/16 169.254.0.0/16 fd00::/8 and fe80::/10, since the
869 RFC standards say these addresses should not be visible on the
870 public internet. Turning on 127.0.0.0/8 would hinder many spam‐
871 blocklists as they use that. Adding ::ffff:0:0/96 stops
872 IPv4-mapped IPv6 addresses from bypassing the filter.
873 private-domain: <domain name>
875 Allow this domain, and all its subdomains to contain private
876 addresses. Give multiple times to allow multiple domain names
877 to contain private addresses. Default is none.
878 unwanted-reply-threshold: <number>
880 If set, a total number of unwanted replies is kept track of in
881 every thread. When it reaches the threshold, a defensive action
882 is taken and a warning is printed to the log. The defensive
883 action is to clear the rrset and message caches, hopefully
884 flushing away any poison. A value of 10 million is suggested.
885 Default is 0 (turned off).
886 do-not-query-address: <IP address>
888 Do not query the given IP address. Can be IP4 or IP6. Append
889 /num to indicate a classless delegation netblock, for example
890 like 10.2.3.4/24 or 2001::11/64.
891 do-not-query-localhost: <yes or no>
893 If yes, localhost is added to the do-not-query-address entries,
894 both IP6 ::1 and IP4 127.0.0.1/8. If no, then localhost can be
895 used to send queries to. Default is yes.
896 prefetch: <yes or no>
898 If yes, message cache elements are prefetched before they expire
899 to keep the cache up to date. Default is no. Turning it on
900 gives about 10 percent more traffic and load on the machine, but
901 popular items do not expire from the cache.
902 prefetch-key: <yes or no>
904 If yes, fetch the DNSKEYs earlier in the validation process,
905 when a DS record is encountered. This lowers the latency of
906 requests. It does use a little more CPU. Also if the cache is
907 set to 0, it is no use. Default is no.
908 deny-any: <yes or no>
910 If yes, deny queries of type ANY with an empty response.
911 Default is no. If disabled, unbound responds with a short list
912 of resource records if some can be found in the cache and makes
913 the upstream type ANY query if there are none.
914 rrset-roundrobin: <yes or no>
916 If yes, Unbound rotates RRSet order in response (the random num‐
917 ber is taken from the query ID, for speed and thread safety).
918 Default is no.
919 minimal-responses: <yes or no>
921 If yes, Unbound doesn't insert authority/additional sections
922 into response messages when those sections are not required.
923 This reduces response size significantly, and may avoid TCP
924 fallback for some responses. This may cause a slight speedup.
925 The default is yes, even though the DNS protocol RFCs mandate
926 these sections, and the additional content could be of use and
927 save roundtrips for clients. Because they are not used, and the
928 saved roundtrips are easier saved with prefetch, whilst this is
929 faster.
930 disable-dnssec-lame-check: <yes or no>
932 If true, disables the DNSSEC lameness check in the iterator.
933 This check sees if RRSIGs are present in the answer, when dnssec
934 is expected, and retries another authority if RRSIGs are unex‐
935 pectedly missing. The validator will insist in RRSIGs for
936 DNSSEC signed domains regardless of this setting, if a trust
937 anchor is loaded.
938 module-config: <"module names">
940 Module configuration, a list of module names separated by spa‐
941 ces, surround the string with quotes (""). The modules can be
942 validator, iterator. Setting this to "iterator" will result in
943 a non-validating server. Setting this to "validator iterator"
944 will turn on DNSSEC validation. The ordering of the modules is
945 important. You must also set trust-anchors for validation to be
946 useful.
947 trust-anchor-file: <filename>
949 File with trusted keys for validation. Both DS and DNSKEY
950 entries can appear in the file. The format of the file is the
951 standard DNS Zone file format. Default is "", or no trust
952 anchor file.
953 auto-trust-anchor-file: <filename>
955 File with trust anchor for one zone, which is tracked with
956 RFC5011 probes. The probes are several times per month, thus
957 the machine must be online frequently. The initial file can be
958 one with contents as described in trust-anchor-file. The file
959 is written to when the anchor is updated, so the unbound user
960 must have write permission. Write permission to the file, but
961 also to the directory it is in (to create a temporary file,
962 which is necessary to deal with filesystem full events), it must
963 also be inside the chroot (if that is used).
964 trust-anchor: <"Resource Record">
966 A DS or DNSKEY RR for a key to use for validation. Multiple
967 entries can be given to specify multiple trusted keys, in addi‐
968 tion to the trust-anchor-files. The resource record is entered
969 in the same format as 'dig' or 'drill' prints them, the same
970 format as in the zone file. Has to be on a single line, with ""
971 around it. A TTL can be specified for ease of cut and paste, but
972 is ignored. A class can be specified, but class IN is default.
973 trusted-keys-file: <filename>
975 File with trusted keys for validation. Specify more than one
976 file with several entries, one file per entry. Like
977 trust-anchor-file but has a different file format. Format is
978 BIND-9 style format, the trusted-keys { name flag proto algo
979 "key"; }; clauses are read. It is possible to use wildcards
980 with this statement, the wildcard is expanded on start and on
981 reload.
982 trust-anchor-signaling: <yes or no>
984 Send RFC8145 key tag query after trust anchor priming. Default
985 is on.
986 root-key-sentinel: <yes or no>
988 Root key trust anchor sentinel. Default is on.
989 dlv-anchor-file: <filename>
991 This option was used during early days DNSSEC deployment when no
992 parent-side DS record registrations were easily available.
993 Nowadays, it is best to have DS records registered with the par‐
994 ent zone (many top level zones are signed). File with trusted
995 keys for DLV (DNSSEC Lookaside Validation). Both DS and DNSKEY
996 entries can be used in the file, in the same format as for
997 trust-anchor-file: statements. Only one DLV can be configured,
998 more would be slow. The DLV configured is used as a root trusted
999 DLV, this means that it is a lookaside for the root. Default is
1000 "", or no dlv anchor file. DLV is going to be decommissioned.
1001 Please do not use it any more.
1002 dlv-anchor: <"Resource Record">
1004 Much like trust-anchor, this is a DLV anchor with the DS or
1005 DNSKEY inline. DLV is going to be decommissioned. Please do
1006 not use it any more.
1007 domain-insecure: <domain name>
1009 Sets domain name to be insecure, DNSSEC chain of trust is
1010 ignored towards the domain name. So a trust anchor above the
1011 domain name can not make the domain secure with a DS record,
1012 such a DS record is then ignored. Also keys from DLV are
1013 ignored for the domain. Can be given multiple times to specify
1014 multiple domains that are treated as if unsigned. If you set
1015 trust anchors for the domain they override this setting (and the
1016 domain is secured).
1017 This can be useful if you want to make sure a trust anchor for
1019 external lookups does not affect an (unsigned) internal domain.
1020 A DS record externally can create validation failures for that
1021 internal domain.
1022 val-override-date: <rrsig-style date spec>
1024 Default is "" or "0", which disables this debugging feature. If
1025 enabled by giving a RRSIG style date, that date is used for ver‐
1026 ifying RRSIG inception and expiration dates, instead of the cur‐
1027 rent date. Do not set this unless you are debugging signature
1028 inception and expiration. The value -1 ignores the date alto‐
1029 gether, useful for some special applications.
1030 val-sig-skew-min: <seconds>
1032 Minimum number of seconds of clock skew to apply to validated
1033 signatures. A value of 10% of the signature lifetime (expira‐
1034 tion - inception) is used, capped by this setting. Default is
1035 3600 (1 hour) which allows for daylight savings differences.
1036 Lower this value for more strict checking of short lived signa‐
1037 tures.
1038 val-sig-skew-max: <seconds>
1040 Maximum number of seconds of clock skew to apply to validated
1041 signatures. A value of 10% of the signature lifetime (expira‐
1042 tion - inception) is used, capped by this setting. Default is
1043 86400 (24 hours) which allows for timezone setting problems in
1044 stable domains. Setting both min and max very low disables the
1045 clock skew allowances. Setting both min and max very high makes
1046 the validator check the signature timestamps less strictly.
1047 val-bogus-ttl: <number>
1049 The time to live for bogus data. This is data that has failed
1050 validation; due to invalid signatures or other checks. The TTL
1051 from that data cannot be trusted, and this value is used
1052 instead. The value is in seconds, default 60. The time interval
1053 prevents repeated revalidation of bogus data.
1054 val-clean-additional: <yes or no>
1056 Instruct the validator to remove data from the additional sec‐
1057 tion of secure messages that are not signed properly. Messages
1058 that are insecure, bogus, indeterminate or unchecked are not
1059 affected. Default is yes. Use this setting to protect the users
1060 that rely on this validator for authentication from potentially
1061 bad data in the additional section.
1062 val-log-level: <number>
1064 Have the validator print validation failures to the log.
1065 Regardless of the verbosity setting. Default is 0, off. At 1,
1066 for every user query that fails a line is printed to the logs.
1067 This way you can monitor what happens with validation. Use a
1068 diagnosis tool, such as dig or drill, to find out why validation
1069 is failing for these queries. At 2, not only the query that
1070 failed is printed but also the reason why unbound thought it was
1071 wrong and which server sent the faulty data.
1072 val-permissive-mode: <yes or no>
1074 Instruct the validator to mark bogus messages as indeterminate.
1075 The security checks are performed, but if the result is bogus
1076 (failed security), the reply is not withheld from the client
1077 with SERVFAIL as usual. The client receives the bogus data. For
1078 messages that are found to be secure the AD bit is set in
1079 replies. Also logging is performed as for full validation. The
1080 default value is "no".
1081 ignore-cd-flag: <yes or no>
1083 Instruct unbound to ignore the CD flag from clients and refuse
1084 to return bogus answers to them. Thus, the CD (Checking Dis‐
1085 abled) flag does not disable checking any more. This is useful
1086 if legacy (w2008) servers that set the CD flag but cannot vali‐
1087 date DNSSEC themselves are the clients, and then unbound pro‐
1088 vides them with DNSSEC protection. The default value is "no".
1089 serve-expired: <yes or no>
1091 If enabled, unbound attempts to serve old responses from cache
1092 with a TTL of 0 in the response without waiting for the actual
1093 resolution to finish. The actual resolution answer ends up in
1094 the cache later on. Default is "no".
1095 serve-expired-ttl: <seconds>
1097 Limit serving of expired responses to configured seconds after
1098 expiration. 0 disables the limit. This option only applies when
1099 serve-expired is enabled. The default is 0.
1100 serve-expired-ttl-reset: <yes or no>
1102 Set the TTL of expired records to the serve-expired-ttl value
1103 after a failed attempt to retrieve the record from upstream.
1104 This makes sure that the expired records will be served as long
1105 as there are queries for it. Default is "no".
1106 val-nsec3-keysize-iterations: <"list of values">
1108 List of keysize and iteration count values, separated by spaces,
1109 surrounded by quotes. Default is "1024 150 2048 500 4096 2500".
1110 This determines the maximum allowed NSEC3 iteration count before
1111 a message is simply marked insecure instead of performing the
1112 many hashing iterations. The list must be in ascending order and
1113 have at least one entry. If you set it to "1024 65535" there is
1114 no restriction to NSEC3 iteration values. This table must be
1115 kept short; a very long list could cause slower operation.
1116 add-holddown: <seconds>
1118 Instruct the auto-trust-anchor-file probe mechanism for RFC5011
1119 autotrust updates to add new trust anchors only after they have
1120 been visible for this time. Default is 30 days as per the RFC.
1121 del-holddown: <seconds>
1123 Instruct the auto-trust-anchor-file probe mechanism for RFC5011
1124 autotrust updates to remove revoked trust anchors after they
1125 have been kept in the revoked list for this long. Default is 30
1126 days as per the RFC.
1127 keep-missing: <seconds>
1129 Instruct the auto-trust-anchor-file probe mechanism for RFC5011
1130 autotrust updates to remove missing trust anchors after they
1131 have been unseen for this long. This cleans up the state file
1132 if the target zone does not perform trust anchor revocation, so
1133 this makes the auto probe mechanism work with zones that perform
1134 regular (non-5011) rollovers. The default is 366 days. The
1135 value 0 does not remove missing anchors, as per the RFC.
1136 permit-small-holddown: <yes or no>
1138 Debug option that allows the autotrust 5011 rollover timers to
1139 assume very small values. Default is no.
1140 key-cache-size: <number>
1142 Number of bytes size of the key cache. Default is 4 megabytes.
1143 A plain number is in bytes, append 'k', 'm' or 'g' for kilo‐
1144 bytes, megabytes or gigabytes (1024*1024 bytes in a megabyte).
1145 key-cache-slabs: <number>
1147 Number of slabs in the key cache. Slabs reduce lock contention
1148 by threads. Must be set to a power of 2. Setting (close) to the
1149 number of cpus is a reasonable guess.
1150 neg-cache-size: <number>
1152 Number of bytes size of the aggressive negative cache. Default
1153 is 1 megabyte. A plain number is in bytes, append 'k', 'm' or
1154 'g' for kilobytes, megabytes or gigabytes (1024*1024 bytes in a
1155 megabyte).
1156 unblock-lan-zones: <yes or no>
1158 Default is disabled. If enabled, then for private address
1159 space, the reverse lookups are no longer filtered. This allows
1160 unbound when running as dns service on a host where it provides
1161 service for that host, to put out all of the queries for the
1162 'lan' upstream. When enabled, only localhost, 127.0.0.1 reverse
1163 and ::1 reverse zones are configured with default local zones.
1164 Disable the option when unbound is running as a (DHCP-) DNS net‐
1165 work resolver for a group of machines, where such lookups should
1166 be filtered (RFC compliance), this also stops potential data
1167 leakage about the local network to the upstream DNS servers.
1168 insecure-lan-zones: <yes or no>
1170 Default is disabled. If enabled, then reverse lookups in pri‐
1171 vate address space are not validated. This is usually required
1172 whenever unblock-lan-zones is used.
1173 local-zone: <zone> <type>
1175 Configure a local zone. The type determines the answer to give
1176 if there is no match from local-data. The types are deny,
1177 refuse, static, transparent, redirect, nodefault, typetranspar‐
1178 ent, inform, inform_deny, always_transparent, always_refuse,
1179 always_nxdomain, noview, and are explained below. After that the
1180 default settings are listed. Use local-data: to enter data into
1181 the local zone. Answers for local zones are authoritative DNS
1182 answers. By default the zones are class IN.
1183 If you need more complicated authoritative data, with referrals,
1185 wildcards, CNAME/DNAME support, or DNSSEC authoritative service,
1186 setup a stub-zone for it as detailed in the stub zone section
1187 below.
1188 deny Do not send an answer, drop the query. If there is a match
1190 from local data, the query is answered.
1191 refuse
1193 Send an error message reply, with rcode REFUSED. If there is
1194 a match from local data, the query is answered.
1195 static
1197 If there is a match from local data, the query is answered.
1198 Otherwise, the query is answered with nodata or nxdomain.
1199 For a negative answer a SOA is included in the answer if
1200 present as local-data for the zone apex domain.
1201 transparent
1203 If there is a match from local data, the query is answered.
1204 Otherwise if the query has a different name, the query is
1205 resolved normally. If the query is for a name given in
1206 localdata but no such type of data is given in localdata,
1207 then a noerror nodata answer is returned. If no local-zone
1208 is given local-data causes a transparent zone to be created
1209 by default.
1210 typetransparent
1212 If there is a match from local data, the query is answered.
1213 If the query is for a different name, or for the same name
1214 but for a different type, the query is resolved normally.
1215 So, similar to transparent but types that are not listed in
1216 local data are resolved normally, so if an A record is in the
1217 local data that does not cause a nodata reply for AAAA
1218 queries.
1219 redirect
1221 The query is answered from the local data for the zone name.
1222 There may be no local data beneath the zone name. This
1223 answers queries for the zone, and all subdomains of the zone
1224 with the local data for the zone. It can be used to redirect
1225 a domain to return a different address record to the end
1226 user, with local-zone: "example.com." redirect and
1227 local-data: "example.com. A 127.0.0.1" queries for www.exam‐
1228 ple.com and www.foo.example.com are redirected, so that users
1229 with web browsers cannot access sites with suffix exam‐
1230 ple.com.
1231 inform
1233 The query is answered normally, same as transparent. The
1234 client IP address (@portnumber) is printed to the logfile.
1235 The log message is: timestamp, unbound-pid, info: zonename
1236 inform IP@port queryname type class. This option can be used
1237 for normal resolution, but machines looking up infected names
1238 are logged, eg. to run antivirus on them.
1239 inform_deny
1241 The query is dropped, like 'deny', and logged, like 'inform'.
1242 Ie. find infected machines without answering the queries.
1243 always_transparent
1245 Like transparent, but ignores local data and resolves nor‐
1246 mally.
1247 always_refuse
1249 Like refuse, but ignores local data and refuses the query.
1250 always_nxdomain
1252 Like static, but ignores local data and returns nxdomain for
1253 the query.
1254 noview
1256 Breaks out of that view and moves towards the global local
1257 zones for answer to the query. If the view first is no,
1258 it'll resolve normally. If view first is enabled, it'll
1259 break perform that step and check the global answers. For
1260 when the view has view specific overrides but some zone has
1261 to be answered from global local zone contents.
1262 nodefault
1264 Used to turn off default contents for AS112 zones. The other
1265 types also turn off default contents for the zone. The 'node‐
1266 fault' option has no other effect than turning off default
1267 contents for the given zone. Use nodefault if you use
1268 exactly that zone, if you want to use a subzone, use trans‐
1269 parent.
1270 The default zones are localhost, reverse 127.0.0.1 and ::1, the onion,
1272 test, invalid and the AS112 zones. The AS112 zones are reverse DNS
1273 zones for private use and reserved IP addresses for which the servers
1274 on the internet cannot provide correct answers. They are configured by
1275 default to give nxdomain (no reverse information) answers. The defaults
1276 can be turned off by specifying your own local-zone of that name, or
1277 using the 'nodefault' type. Below is a list of the default zone con‐
1278 tents.
1279 localhost
1281 The IP4 and IP6 localhost information is given. NS and SOA
1282 records are provided for completeness and to satisfy some DNS
1283 update tools. Default content:
1284 local-zone: "localhost." redirect
1285 local-data: "localhost. 10800 IN NS localhost."
1286 local-data: "localhost. 10800 IN
1287 SOA localhost. nobody.invalid. 1 3600 1200 604800 10800"
1288 local-data: "localhost. 10800 IN A 127.0.0.1"
1289 local-data: "localhost. 10800 IN AAAA ::1"
1290 reverse IPv4 loopback
1292 Default content:
1293 local-zone: "127.in-addr.arpa." static
1294 local-data: "127.in-addr.arpa. 10800 IN NS localhost."
1295 local-data: "127.in-addr.arpa. 10800 IN
1296 SOA localhost. nobody.invalid. 1 3600 1200 604800 10800"
1297 local-data: "1.0.0.127.in-addr.arpa. 10800 IN
1298 PTR localhost."
1299 reverse IPv6 loopback
1301 Default content:
1302 local-zone: "1.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.
1303 0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.ip6.arpa." static
1304 local-data: "1.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.
1305 0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.ip6.arpa. 10800 IN
1306 NS localhost."
1307 local-data: "1.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.
1308 0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.ip6.arpa. 10800 IN
1309 SOA localhost. nobody.invalid. 1 3600 1200 604800 10800"
1310 local-data: "1.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.
1311 0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.ip6.arpa. 10800 IN
1312 PTR localhost."
1313 onion (RFC 7686)
1315 Default content:
1316 local-zone: "onion." static
1317 local-data: "onion. 10800 IN NS localhost."
1318 local-data: "onion. 10800 IN
1319 SOA localhost. nobody.invalid. 1 3600 1200 604800 10800"
1320 test (RFC 2606)
1322 Default content:
1323 local-zone: "test." static
1324 local-data: "test. 10800 IN NS localhost."
1325 local-data: "test. 10800 IN
1326 SOA localhost. nobody.invalid. 1 3600 1200 604800 10800"
1327 invalid (RFC 2606)
1329 Default content:
1330 local-zone: "invalid." static
1331 local-data: "invalid. 10800 IN NS localhost."
1332 local-data: "invalid. 10800 IN
1333 SOA localhost. nobody.invalid. 1 3600 1200 604800 10800"
1334 reverse RFC1918 local use zones
1336 Reverse data for zones 10.in-addr.arpa, 16.172.in-addr.arpa
1337 to 31.172.in-addr.arpa, 168.192.in-addr.arpa. The
1338 local-zone: is set static and as local-data: SOA and NS
1339 records are provided.
1340 reverse RFC3330 IP4 this, link-local, testnet and broadcast
1342 Reverse data for zones 0.in-addr.arpa, 254.169.in-addr.arpa,
1343 2.0.192.in-addr.arpa (TEST NET 1), 100.51.198.in-addr.arpa
1344 (TEST NET 2), 113.0.203.in-addr.arpa (TEST NET 3),
1345 255.255.255.255.in-addr.arpa. And from 64.100.in-addr.arpa
1346 to 127.100.in-addr.arpa (Shared Address Space).
1347 reverse RFC4291 IP6 unspecified
1349 Reverse data for zone
1350 0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.
1351 0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.ip6.arpa.
1352 reverse RFC4193 IPv6 Locally Assigned Local Addresses
1354 Reverse data for zone D.F.ip6.arpa.
1355 reverse RFC4291 IPv6 Link Local Addresses
1357 Reverse data for zones 8.E.F.ip6.arpa to B.E.F.ip6.arpa.
1358 reverse IPv6 Example Prefix
1360 Reverse data for zone 8.B.D.0.1.0.0.2.ip6.arpa. This zone is
1361 used for tutorials and examples. You can remove the block on
1362 this zone with:
1363 local-zone: 8.B.D.0.1.0.0.2.ip6.arpa. nodefault
1364 You can also selectively unblock a part of the zone by making
1365 that part transparent with a local-zone statement. This also
1366 works with the other default zones.
1367 local-data: "<resource record string>"
1369 Configure local data, which is served in reply to queries for it.
1370 The query has to match exactly unless you configure the local-zone
1371 as redirect. If not matched exactly, the local-zone type deter‐
1372 mines further processing. If local-data is configured that is not
1373 a subdomain of a local-zone, a transparent local-zone is config‐
1374 ured. For record types such as TXT, use single quotes, as in
1375 local-data: 'example. TXT "text"'.
1376 If you need more complicated authoritative data, with referrals,
1378 wildcards, CNAME/DNAME support, or DNSSEC authoritative service,
1379 setup a stub-zone for it as detailed in the stub zone section
1380 below.
1381 local-data-ptr: "IPaddr name"
1383 Configure local data shorthand for a PTR record with the reversed
1384 IPv4 or IPv6 address and the host name. For example "192.0.2.4
1385 www.example.com". TTL can be inserted like this: "2001:DB8::4
1386 7200 www.example.com"
1387 local-zone-tag: <zone> <"list of tags">
1389 Assign tags to localzones. Tagged localzones will only be applied
1390 when the used access-control element has a matching tag. Tags must
1391 be defined in define-tags. Enclose list of tags in quotes ("")
1392 and put spaces between tags.
1393 local-zone-override: <zone> <IP netblock> <type>
1395 Override the localzone type for queries from addresses matching
1396 netblock. Use this localzone type, regardless the type configured
1397 for the local-zone (both tagged and untagged) and regardless the
1398 type configured using access-control-tag-action.
1399 ratelimit: <number or 0>
1401 Enable ratelimiting of queries sent to nameserver for performing
1402 recursion. If 0, the default, it is disabled. This option is
1403 experimental at this time. The ratelimit is in queries per second
1404 that are allowed. More queries are turned away with an error
1405 (servfail). This stops recursive floods, eg. random query names,
1406 but not spoofed reflection floods. Cached responses are not rate‐
1407 limited by this setting. The zone of the query is determined by
1408 examining the nameservers for it, the zone name is used to keep
1409 track of the rate. For example, 1000 may be a suitable value to
1410 stop the server from being overloaded with random names, and keeps
1411 unbound from sending traffic to the nameservers for those zones.
1412 ratelimit-size: <memory size>
1414 Give the size of the data structure in which the current ongoing
1415 rates are kept track in. Default 4m. In bytes or use m(mega),
1416 k(kilo), g(giga). The ratelimit structure is small, so this data
1417 structure likely does not need to be large.
1418 ratelimit-slabs: <number>
1420 Give power of 2 number of slabs, this is used to reduce lock con‐
1421 tention in the ratelimit tracking data structure. Close to the
1422 number of cpus is a fairly good setting.
1423 ratelimit-factor: <number>
1425 Set the amount of queries to rate limit when the limit is
1426 exceeded. If set to 0, all queries are dropped for domains where
1427 the limit is exceeded. If set to another value, 1 in that number
1428 is allowed through to complete. Default is 10, allowing 1/10
1429 traffic to flow normally. This can make ordinary queries complete
1430 (if repeatedly queried for), and enter the cache, whilst also mit‐
1431 igating the traffic flow by the factor given.
1432 ratelimit-for-domain: <domain> <number qps or 0>
1434 Override the global ratelimit for an exact match domain name with
1435 the listed number. You can give this for any number of names.
1436 For example, for a top-level-domain you may want to have a higher
1437 limit than other names. A value of 0 will disable ratelimiting
1438 for that domain.
1439 ratelimit-below-domain: <domain> <number qps or 0>
1441 Override the global ratelimit for a domain name that ends in this
1442 name. You can give this multiple times, it then describes differ‐
1443 ent settings in different parts of the namespace. The closest
1444 matching suffix is used to determine the qps limit. The rate for
1445 the exact matching domain name is not changed, use rate‐
1446 limit-for-domain to set that, you might want to use different set‐
1447 tings for a top-level-domain and subdomains. A value of 0 will
1448 disable ratelimiting for domain names that end in this name.
1449 ip-ratelimit: <number or 0>
1451 Enable global ratelimiting of queries accepted per ip address. If
1452 0, the default, it is disabled. This option is experimental at
1453 this time. The ratelimit is in queries per second that are
1454 allowed. More queries are completely dropped and will not receive
1455 a reply, SERVFAIL or otherwise. IP ratelimiting happens before
1456 looking in the cache. This may be useful for mitigating amplifica‐
1457 tion attacks.
1458 ip-ratelimit-size: <memory size>
1460 Give the size of the data structure in which the current ongoing
1461 rates are kept track in. Default 4m. In bytes or use m(mega),
1462 k(kilo), g(giga). The ip ratelimit structure is small, so this
1463 data structure likely does not need to be large.
1464 ip-ratelimit-slabs: <number>
1466 Give power of 2 number of slabs, this is used to reduce lock con‐
1467 tention in the ip ratelimit tracking data structure. Close to the
1468 number of cpus is a fairly good setting.
1469 ip-ratelimit-factor: <number>
1471 Set the amount of queries to rate limit when the limit is
1472 exceeded. If set to 0, all queries are dropped for addresses
1473 where the limit is exceeded. If set to another value, 1 in that
1474 number is allowed through to complete. Default is 10, allowing
1475 1/10 traffic to flow normally. This can make ordinary queries
1476 complete (if repeatedly queried for), and enter the cache, whilst
1477 also mitigating the traffic flow by the factor given.
1478 fast-server-permil: <number>
1480 Specify how many times out of 1000 to pick from the set of fastest
1481 servers. 0 turns the feature off. A value of 900 would pick from
1482 the fastest servers 90 percent of the time, and would perform nor‐
1483 mal exploration of random servers for the remaining time. When
1484 prefetch is enabled (or serve-expired), such prefetches are not
1485 sped up, because there is no one waiting for it, and it presents a
1486 good moment to perform server exploration. The fast-server-num
1487 option can be used to specify the size of the fastest servers set.
1488 The default for fast-server-permil is 0.
1489 fast-server-num: <number>
1491 Set the number of servers that should be used for fast server
1492 selection. Only use the fastest specified number of servers with
1493 the fast-server-permil option, that turns this on or off. The
1494 default is to use the fastest 3 servers.
1495 Remote Control Options
1497 In the remote-control: clause are the declarations for the remote con‐
1498 trol facility. If this is enabled, the unbound-control(8) utility can
1499 be used to send commands to the running unbound server. The server
1500 uses these clauses to setup TLSv1 security for the connection. The
1501 unbound-control(8) utility also reads the remote-control section for
1502 options. To setup the correct self-signed certificates use the
1503 unbound-control-setup(8) utility.
1504 control-enable: <yes or no>
1506 The option is used to enable remote control, default is "no". If
1507 turned off, the server does not listen for control commands.
1508 control-interface: <ip address or path>
1510 Give IPv4 or IPv6 addresses or local socket path to listen on for
1511 control commands. By default localhost (127.0.0.1 and ::1) is
1512 listened to. Use 0.0.0.0 and ::0 to listen to all interfaces. If
1513 you change this and permissions have been dropped, you must
1514 restart the server for the change to take effect.
1515 If you set it to an absolute path, a local socket is used. The
1517 local socket does not use the certificates and keys, so those
1518 files need not be present. To restrict access, unbound sets per‐
1519 missions on the file to the user and group that is configured, the
1520 access bits are set to allow the group members to access the con‐
1521 trol socket file. Put users that need to access the socket in the
1522 that group. To restrict access further, create a directory to put
1523 the control socket in and restrict access to that directory.
1524 control-port: <port number>
1526 The port number to listen on for IPv4 or IPv6 control interfaces,
1527 default is 8953. If you change this and permissions have been
1528 dropped, you must restart the server for the change to take
1529 effect.
1530 control-use-cert: <yes or no>
1532 For localhost control-interface you can disable the use of TLS by
1533 setting this option to "no", default is "yes". For local sockets,
1534 TLS is disabled and the value of this option is ignored.
1535 server-key-file: <private key file>
1537 Path to the server private key, by default unbound_server.key.
1538 This file is generated by the unbound-control-setup utility. This
1539 file is used by the unbound server, but not by unbound-control.
1540 server-cert-file: <certificate file.pem>
1542 Path to the server self signed certificate, by default
1543 unbound_server.pem. This file is generated by the unbound-con‐
1544 trol-setup utility. This file is used by the unbound server, and
1545 also by unbound-control.
1546 control-key-file: <private key file>
1548 Path to the control client private key, by default unbound_con‐
1549 trol.key. This file is generated by the unbound-control-setup
1550 utility. This file is used by unbound-control.
1551 control-cert-file: <certificate file.pem>
1553 Path to the control client certificate, by default unbound_con‐
1554 trol.pem. This certificate has to be signed with the server cer‐
1555 tificate. This file is generated by the unbound-control-setup
1556 utility. This file is used by unbound-control.
1557 Stub Zone Options
1559 There may be multiple stub-zone: clauses. Each with a name: and zero or
1560 more hostnames or IP addresses. For the stub zone this list of name‐
1561 servers is used. Class IN is assumed. The servers should be authority
1562 servers, not recursors; unbound performs the recursive processing
1563 itself for stub zones.
1564 The stub zone can be used to configure authoritative data to be used by
1566 the resolver that cannot be accessed using the public internet servers.
1567 This is useful for company-local data or private zones. Setup an
1568 authoritative server on a different host (or different port). Enter a
1569 config entry for unbound with stub-addr: <ip address of host[@port]>.
1570 The unbound resolver can then access the data, without referring to the
1571 public internet for it.
1572 This setup allows DNSSEC signed zones to be served by that authorita‐
1574 tive server, in which case a trusted key entry with the public key can
1575 be put in config, so that unbound can validate the data and set the AD
1576 bit on replies for the private zone (authoritative servers do not set
1577 the AD bit). This setup makes unbound capable of answering queries for
1578 the private zone, and can even set the AD bit ('authentic'), but the AA
1579 ('authoritative') bit is not set on these replies.
1580 Consider adding server: statements for domain-insecure: and for
1582 local-zone: name nodefault for the zone if it is a locally served zone.
1583 The insecure clause stops DNSSEC from invalidating the zone. The local
1584 zone nodefault (or transparent) clause makes the (reverse-) zone bypass
1585 unbound's filtering of RFC1918 zones.
1586 name: <domain name>
1588 Name of the stub zone.
1589 stub-host: <domain name>
1591 Name of stub zone nameserver. Is itself resolved before it is
1592 used.
1593 stub-addr: <IP address>
1595 IP address of stub zone nameserver. Can be IP 4 or IP 6. To use
1596 a nondefault port for DNS communication append '@' with the port
1597 number.
1598 stub-prime: <yes or no>
1600 This option is by default no. If enabled it performs NS set
1601 priming, which is similar to root hints, where it starts using
1602 the list of nameservers currently published by the zone. Thus,
1603 if the hint list is slightly outdated, the resolver picks up a
1604 correct list online.
1605 stub-first: <yes or no>
1607 If enabled, a query is attempted without the stub clause if it
1608 fails. The data could not be retrieved and would have caused
1609 SERVFAIL because the servers are unreachable, instead it is
1610 tried without this clause. The default is no.
1611 stub-tls-upstream: <yes or no>
1613 Enabled or disable whether the queries to this stub use TLS for
1614 transport. Default is no.
1615 stub-ssl-upstream: <yes or no>
1617 Alternate syntax for stub-tls-upstream.
1618 stub-no-cache: <yes or no>
1620 Default is no. If enabled, data inside the stub is not cached.
1621 This is useful when you want immediate changes to be visible.
1622 Forward Zone Options
1624 There may be multiple forward-zone: clauses. Each with a name: and zero
1625 or more hostnames or IP addresses. For the forward zone this list of
1626 nameservers is used to forward the queries to. The servers listed as
1627 forward-host: and forward-addr: have to handle further recursion for
1628 the query. Thus, those servers are not authority servers, but are
1629 (just like unbound is) recursive servers too; unbound does not perform
1630 recursion itself for the forward zone, it lets the remote server do it.
1631 Class IN is assumed. CNAMEs are chased by unbound itself, asking the
1632 remote server for every name in the indirection chain, to protect the
1633 local cache from illegal indirect referenced items. A forward-zone
1634 entry with name "." and a forward-addr target will forward all queries
1635 to that other server (unless it can answer from the cache).
1636 name: <domain name>
1638 Name of the forward zone.
1639 forward-host: <domain name>
1641 Name of server to forward to. Is itself resolved before it is
1642 used.
1643 forward-addr: <IP address>
1645 IP address of server to forward to. Can be IP 4 or IP 6. To use
1646 a nondefault port for DNS communication append '@' with the port
1647 number. If tls is enabled, then you can append a '#' and a
1648 name, then it'll check the tls authentication certificates with
1649 that name. If you combine the '@' and '#', the '@' comes first.
1650 At high verbosity it logs the TLS certificate, with TLS enabled.
1652 If you leave out the '#' and auth name from the forward-addr,
1653 any name is accepted. The cert must also match a CA from the
1654 tls-cert-bundle. The cert name match code needs OpenSSL 1.1.0
1655 or later to be enabled.
1656 forward-first: <yes or no>
1658 If enabled, a query is attempted without the forward clause if
1659 it fails. The data could not be retrieved and would have caused
1660 SERVFAIL because the servers are unreachable, instead it is
1661 tried without this clause. The default is no.
1662 forward-tls-upstream: <yes or no>
1664 Enabled or disable whether the queries to this forwarder use TLS
1665 for transport. Default is no. If you enable this, also config‐
1666 ure a tls-cert-bundle or use tls-win-cert to load CA certs, oth‐
1667 erwise the connections cannot be authenticated.
1668 forward-ssl-upstream: <yes or no>
1670 Alternate syntax for forward-tls-upstream.
1671 forward-no-cache: <yes or no>
1673 Default is no. If enabled, data inside the forward is not
1674 cached. This is useful when you want immediate changes to be
1675 visible.
1676 Authority Zone Options
1678 Authority zones are configured with auth-zone:, and each one must have
1679 a name:. There can be multiple ones, by listing multiple auth-zone
1680 clauses, each with a different name, pertaining to that part of the
1681 namespace. The authority zone with the name closest to the name looked
1682 up is used. Authority zones are processed after local-zones and before
1683 cache (for-downstream: yes), and when used in this manner make unbound
1684 respond like an authority server. Authority zones are also processed
1685 after cache, just before going to the network to fetch information for
1686 recursion (for-upstream: yes), and when used in this manner provide a
1687 local copy of an authority server that speeds up lookups of that data.
1688 Authority zones can be read from zonefile. And can be kept updated via
1690 AXFR and IXFR. After update the zonefile is rewritten. The update
1691 mechanism uses the SOA timer values and performs SOA UDP queries to
1692 detect zone changes.
1693 If the update fetch fails, the timers in the SOA record are used to
1695 time another fetch attempt. Until the SOA expiry timer is reached.
1696 Then the zone is expired. When a zone is expired, queries are SERV‐
1697 FAIL, and any new serial number is accepted from the master (even if
1698 older), and if fallback is enabled, the fallback activates to fetch
1699 from the upstream instead of the SERVFAIL.
1700 name: <zone name>
1702 Name of the authority zone.
1703 master: <IP address or host name>
1705 Where to download a copy of the zone from, with AXFR and IXFR.
1706 Multiple masters can be specified. They are all tried if one
1707 fails.
1708 url: <url to zonefile>
1710 Where to download a zonefile for the zone. With http or https.
1711 An example for the url is "http://www.example.com/exam‐
1712 ple.org.zone". Multiple url statements can be given, they are
1713 tried in turn. If only urls are given the SOA refresh timer is
1714 used to wait for making new downloads. If also masters are
1715 listed, the masters are first probed with UDP SOA queries to see
1716 if the SOA serial number has changed, reducing the number of
1717 downloads. If none of the urls work, the masters are tried with
1718 IXFR and AXFR. For https, the tls-cert-bundle and the hostname
1719 from the url are used to authenticate the connection.
1720 allow-notify: <IP address or host name or netblockIP/prefix>
1722 With allow-notify you can specify additional sources of noti‐
1723 fies. When notified, the server attempts to first probe and
1724 then zone transfer. If the notify is from a master, it first
1725 attempts that master. Otherwise other masters are attempted.
1726 If there are no masters, but only urls, the file is downloaded
1727 when notified. The masters from master: statements are allowed
1728 notify by default.
1729 fallback-enabled: <yes or no>
1731 Default no. If enabled, unbound falls back to querying the
1732 internet as a resolver for this zone when lookups fail. For
1733 example for DNSSEC validation failures.
1734 for-downstream: <yes or no>
1736 Default yes. If enabled, unbound serves authority responses to
1737 downstream clients for this zone. This option makes unbound
1738 behave, for the queries with names in this zone, like one of the
1739 authority servers for that zone. Turn it off if you want
1740 unbound to provide recursion for the zone but have a local copy
1741 of zone data. If for-downstream is no and for-upstream is yes,
1742 then unbound will DNSSEC validate the contents of the zone
1743 before serving the zone contents to clients and store validation
1744 results in the cache.
1745 for-upstream: <yes or no>
1747 Default yes. If enabled, unbound fetches data from this data
1748 collection for answering recursion queries. Instead of sending
1749 queries over the internet to the authority servers for this
1750 zone, it'll fetch the data directly from the zone data. Turn it
1751 on when you want unbound to provide recursion for downstream
1752 clients, and use the zone data as a local copy to speed up
1753 lookups.
1754 zonefile: <filename>
1756 The filename where the zone is stored. If not given then no
1757 zonefile is used. If the file does not exist or is empty,
1758 unbound will attempt to fetch zone data (eg. from the master
1759 servers).
1760 View Options
1762 There may be multiple view: clauses. Each with a name: and zero or more
1763 local-zone and local-data elements. View can be mapped to requests by
1764 specifying the view name in an access-control-view element. Options
1765 from matching views will override global options. Global options will
1766 be used if no matching view is found, or when the matching view does
1767 not have the option specified.
1768 name: <view name>
1770 Name of the view. Must be unique. This name is used in
1771 access-control-view elements.
1772 local-zone: <zone> <type>
1774 View specific local-zone elements. Has the same types and behav‐
1775 iour as the global local-zone elements. When there is at least
1776 one local-zone specified and view-first is no, the default
1777 local-zones will be added to this view. Defaults can be dis‐
1778 abled using the nodefault type. When view-first is yes or when a
1779 view does not have a local-zone, the global local-zone will be
1780 used including it's default zones.
1781 local-data: "<resource record string>"
1783 View specific local-data elements. Has the same behaviour as the
1784 global local-data elements.
1785 local-data-ptr: "IPaddr name"
1787 View specific local-data-ptr elements. Has the same behaviour as
1788 the global local-data-ptr elements.
1789 view-first: <yes or no>
1791 If enabled, it attempts to use the global local-zone and
1792 local-data if there is no match in the view specific options.
1793 The default is no.
1794 Python Module Options
1796 The python: clause gives the settings for the python(1) script module.
1797 This module acts like the iterator and validator modules do, on queries
1798 and answers. To enable the script module it has to be compiled into
1799 the daemon, and the word "python" has to be put in the module-config:
1800 option (usually first, or between the validator and iterator).
1801 If the chroot: option is enabled, you should make sure Python's library
1803 directory structure is bind mounted in the new root environment, see
1804 mount(8). Also the python-script: path should be specified as an abso‐
1805 lute path relative to the new root, or as a relative path to the work‐
1806 ing directory.
1807 python-script: <python file>
1809 The script file to load.
1810 DNS64 Module Options
1812 The dns64 module must be configured in the module-config: "dns64 val‐
1813 idator iterator" directive and be compiled into the daemon to be
1814 enabled. These settings go in the server: section.
1815 dns64-prefix: <IPv6 prefix>
1817 This sets the DNS64 prefix to use to synthesize AAAA records
1818 with. It must be /96 or shorter. The default prefix is
1819 64:ff9b::/96.
1820 dns64-synthall: <yes or no>
1822 Debug option, default no. If enabled, synthesize all AAAA
1823 records despite the presence of actual AAAA records.
1824 dns64-ignore-aaaa: <name>
1826 List domain for which the AAAA records are ignored and the A
1827 record is used by dns64 processing instead. Can be entered mul‐
1828 tiple times, list a new domain for which it applies, one per
1829 line. Applies also to names underneath the name given.
1830 DNSCrypt Options
1832 The dnscrypt: clause gives the settings of the dnscrypt channel. While
1833 those options are available, they are only meaningful if unbound was
1834 compiled with --enable-dnscrypt. Currently certificate and secret/pub‐
1835 lic keys cannot be generated by unbound. You can use dnscrypt-wrapper
1836 to generate those: https://github.com/cofyc/dnscrypt-wrapper/blob/mas‐
1837 ter/README.md#usage
1838 dnscrypt-enable: <yes or no>
1840 Whether or not the dnscrypt config should be enabled. You may
1841 define configuration but not activate it. The default is no.
1842 dnscrypt-port: <port number>
1844 On which port should dnscrypt should be activated. Note that you
1845 should have a matching interface option defined in the server
1846 section for this port.
1847 dnscrypt-provider: <provider name>
1849 The provider name to use to distribute certificates. This is of
1850 the form: 2.dnscrypt-cert.example.com.. The name MUST end with a
1851 dot.
1852 dnscrypt-secret-key: <path to secret key file>
1854 Path to the time limited secret key file. This option may be
1855 specified multiple times.
1856 dnscrypt-provider-cert: <path to cert file>
1858 Path to the certificate related to the dnscrypt-secret-keys.
1859 This option may be specified multiple times.
1860 dnscrypt-provider-cert-rotated: <path to cert file>
1862 Path to a certificate that we should be able to serve existing
1863 connection from but do not want to advertise over
1864 dnscrypt-provider's TXT record certs distribution. A typical
1865 use case is when rotating certificates, existing clients may
1866 still use the client magic from the old cert in their queries
1867 until they fetch and update the new cert. Likewise, it would
1868 allow one to prime the new cert/key without distributing the new
1869 cert yet, this can be useful when using a network of servers
1870 using anycast and on which the configuration may not get updated
1871 at the exact same time. By priming the cert, the servers can
1872 handle both old and new certs traffic while distributing only
1873 one. This option may be specified multiple times.
1874 dnscrypt-shared-secret-cache-size: <memory size>
1876 Give the size of the data structure in which the shared secret
1877 keys are kept in. Default 4m. In bytes or use m(mega),
1878 k(kilo), g(giga). The shared secret cache is used when a same
1879 client is making multiple queries using the same public key. It
1880 saves a substantial amount of CPU.
1881 dnscrypt-shared-secret-cache-slabs: <number>
1883 Give power of 2 number of slabs, this is used to reduce lock
1884 contention in the dnscrypt shared secrets cache. Close to the
1885 number of cpus is a fairly good setting.
1886 dnscrypt-nonce-cache-size: <memory size>
1888 Give the size of the data structure in which the client nonces
1889 are kept in. Default 4m. In bytes or use m(mega), k(kilo),
1890 g(giga). The nonce cache is used to prevent dnscrypt message
1891 replaying. Client nonce should be unique for any pair of client
1892 pk/server sk.
1893 dnscrypt-nonce-cache-slabs: <number>
1895 Give power of 2 number of slabs, this is used to reduce lock
1896 contention in the dnscrypt nonce cache. Close to the number of
1897 cpus is a fairly good setting.
1898 EDNS Client Subnet Module Options
1900 The ECS module must be configured in the module-config: "subnetcache
1901 validator iterator" directive and be compiled into the daemon to be
1902 enabled. These settings go in the server: section.
1903 If the destination address is whitelisted with Unbound will add the
1905 EDNS0 option to the query containing the relevant part of the client's
1906 address. When an answer contains the ECS option the response and the
1907 option are placed in a specialized cache. If the authority indicated no
1908 support, the response is stored in the regular cache.
1909 Additionally, when a client includes the option in its queries, Unbound
1911 will forward the option to the authority if present in the whitelist,
1912 or client-subnet-always-forward is set to yes. In this case the lookup
1913 in the regular cache is skipped.
1914 The maximum size of the ECS cache is controlled by 'msg-cache-size' in
1916 the configuration file. On top of that, for each query only 100 differ‐
1917 ent subnets are allowed to be stored for each address family. Exceeding
1918 that number, older entries will be purged from cache.
1919 send-client-subnet: <IP address>
1921 Send client source address to this authority. Append /num to
1922 indicate a classless delegation netblock, for example like
1923 10.2.3.4/24 or 2001::11/64. Can be given multiple times. Author‐
1924 ities not listed will not receive edns-subnet information,
1925 unless domain in query is specified in client-subnet-zone.
1926 client-subnet-zone: <domain>
1928 Send client source address in queries for this domain and its
1929 subdomains. Can be given multiple times. Zones not listed will
1930 not receive edns-subnet information, unless hosted by authority
1931 specified in send-client-subnet.
1932 client-subnet-always-forward: <yes or no>
1934 Specify whether the ECS whitelist check (configured using
1935 send-client-subnet) is applied for all queries, even if the
1936 triggering query contains an ECS record, or only for queries for
1937 which the ECS record is generated using the querier address (and
1938 therefore did not contain ECS data in the client query). If
1939 enabled, the whitelist check is skipped when the client query
1940 contains an ECS record. Default is no.
1941 max-client-subnet-ipv6: <number>
1943 Specifies the maximum prefix length of the client source address
1944 we are willing to expose to third parties for IPv6. Defaults to
1945 56.
1946 max-client-subnet-ipv4: <number>
1948 Specifies the maximum prefix length of the client source address
1949 we are willing to expose to third parties for IPv4. Defaults to
1950 24.
1951 min-client-subnet-ipv6: <number>
1953 Specifies the minimum prefix length of the IPv6 source mask we
1954 are willing to accept in queries. Shorter source masks result in
1955 REFUSED answers. Source mask of 0 is always accepted. Default is
1956 0.
1957 min-client-subnet-ipv4: <number>
1959 Specifies the minimum prefix length of the IPv4 source mask we
1960 are willing to accept in queries. Shorter source masks result in
1961 REFUSED answers. Source mask of 0 is always accepted. Default is
1962 0.
1963 max-ecs-tree-size-ipv4: <number>
1965 Specifies the maximum number of subnets ECS answers kept in the
1966 ECS radix tree. This number applies for each qname/qclass/qtype
1967 tuple. Defaults to 100.
1968 max-ecs-tree-size-ipv6: <number>
1970 Specifies the maximum number of subnets ECS answers kept in the
1971 ECS radix tree. This number applies for each qname/qclass/qtype
1972 tuple. Defaults to 100.
1973 Opportunistic IPsec Support Module Options
1975 The IPsec module must be configured in the module-config: "ipsecmod
1976 validator iterator" directive and be compiled into the daemon to be
1977 enabled. These settings go in the server: section.
1978 When unbound receives an A/AAAA query that is not in the cache and
1980 finds a valid answer, it will withhold returning the answer and instead
1981 will generate an IPSECKEY subquery for the same domain name. If an
1982 answer was found, unbound will call an external hook passing the fol‐
1983 lowing arguments:
1984 QNAME
1986 Domain name of the A/AAAA and IPSECKEY query. In string for‐
1987 mat.
1988 IPSECKEY TTL
1990 TTL of the IPSECKEY RRset.
1991 A/AAAA
1993 String of space separated IP addresses present in the A/AAAA
1994 RRset. The IP addresses are in string format.
1995 IPSECKEY
1997 String of space separated IPSECKEY RDATA present in the
1998 IPSECKEY RRset. The IPSECKEY RDATA are in DNS presentation
1999 format.
2000 The A/AAAA answer is then cached and returned to the client. If the
2002 external hook was called the TTL changes to ensure it doesn't surpass
2003 ipsecmod-max-ttl.
2004 The same procedure is also followed when prefetch: is used, but the
2006 A/AAAA answer is given to the client before the hook is called. ipsec‐
2007 mod-max-ttl ensures that the A/AAAA answer given from cache is still
2008 relevant for opportunistic IPsec.
2009 ipsecmod-enabled: <yes or no>
2011 Specifies whether the IPsec module is enabled or not. The IPsec
2012 module still needs to be defined in the module-config: direc‐
2013 tive. This option facilitates turning on/off the module without
2014 restarting/reloading unbound. Defaults to yes.
2015 ipsecmod-hook: <filename>
2017 Specifies the external hook that unbound will call with sys‐
2018 tem(3). The file can be specified as an absolute/relative path.
2019 The file needs the proper permissions to be able to be executed
2020 by the same user that runs unbound. It must be present when the
2021 IPsec module is defined in the module-config: directive.
2022 ipsecmod-strict: <yes or no>
2024 If enabled unbound requires the external hook to return a suc‐
2025 cess value of 0. Failing to do so unbound will reply with SERV‐
2026 FAIL. The A/AAAA answer will also not be cached. Defaults to
2027 no.
2028 ipsecmod-max-ttl: <seconds>
2030 Time to live maximum for A/AAAA cached records after calling the
2031 external hook. Defaults to 3600.
2032 ipsecmod-ignore-bogus: <yes or no>
2034 Specifies the behaviour of unbound when the IPSECKEY answer is
2035 bogus. If set to yes, the hook will be called and the A/AAAA
2036 answer will be returned to the client. If set to no, the hook
2037 will not be called and the answer to the A/AAAA query will be
2038 SERVFAIL. Mainly used for testing. Defaults to no.
2039 ipsecmod-whitelist: <domain>
2041 Whitelist the domain so that the module logic will be executed.
2042 Can be given multiple times, for different domains. If the
2043 option is not specified, all domains are treated as being
2044 whitelisted (default).
2045 Cache DB Module Options
2047 The Cache DB module must be configured in the module-config: "validator
2048 cachedb iterator" directive and be compiled into the daemon with
2049 --enable-cachedb. If this module is enabled and configured, the speci‐
2050 fied backend database works as a second level cache: When Unbound can‐
2051 not find an answer to a query in its built-in in-memory cache, it con‐
2052 sults the specified backend. If it finds a valid answer in the back‐
2053 end, Unbound uses it to respond to the query without performing itera‐
2054 tive DNS resolution. If Unbound cannot even find an answer in the
2055 backend, it resolves the query as usual, and stores the answer in the
2056 backend.
2057 If Unbound was built with --with-libhiredis on a system that has
2059 installed the hiredis C client library of Redis, then the "redis" back‐
2060 end can be used. This backend communicates with the specified Redis
2061 server over a TCP connection to store and retrieve cache data. It can
2062 be used as a persistent and/or shared cache backend. It should be
2063 noted that Unbound never removes data stored in the Redis server, even
2064 if some data have expired in terms of DNS TTL or the Redis server has
2065 cached too much data; if necessary the Redis server must be configured
2066 to limit the cache size, preferably with some kind of least-recently-
2067 used eviction policy. This backend uses synchronous communication with
2068 the Redis server based on the assumption that the communication is sta‐
2069 ble and sufficiently fast. The thread waiting for a response from the
2070 Redis server cannot handle other DNS queries. Although the backend has
2071 the ability to reconnect to the server when the connection is closed
2072 unexpectedly and there is a configurable timeout in case the server is
2073 overly slow or hangs up, these cases are assumed to be very rare. If
2074 connection close or timeout happens too often, Unbound will be effec‐
2075 tively unusable with this backend. It's the administrator's responsi‐
2076 bility to make the assumption hold.
2077 The cachedb: clause gives custom settings of the cache DB module.
2079 backend: <backend name>
2081 Specify the backend database name. The default database is the
2082 in-memory backend named "testframe", which, as the name sug‐
2083 gests, is not of any practical use. Depending on the build-time
2084 configuration, "redis" backend may also be used as described
2085 above.
2086 secret-seed: <"secret string">
2088 Specify a seed to calculate a hash value from query information.
2089 This value will be used as the key of the corresponding answer
2090 for the backend database and can be customized if the hash
2091 should not be predictable operationally. If the backend data‐
2092 base is shared by multiple Unbound instances, all instances must
2093 use the same secret seed. This option defaults to "default".
2094 The following cachedb otions are specific to the redis backend.
2096 redis-server-host: <server address or name>
2098 The IP (either v6 or v4) address or domain name of the Redis
2099 server. In general an IP address should be specified as other‐
2100 wise Unbound will have to resolve the name of the server every
2101 time it establishes a connection to the server. This option
2102 defaults to "127.0.0.1".
2103 redis-server-port: <port number>
2105 The TCP port number of the Redis server. This option defaults
2106 to 6379.
2107 redis-timeout: <msec>
2109 The period until when Unbound waits for a response from the
2110 Redis sever. If this timeout expires Unbound closes the connec‐
2111 tion, treats it as if the Redis server does not have the
2112 requested data, and will try to re-establish a new connection
2113 later. This option defaults to 100 milliseconds.
2114
2116 In the example config settings below memory usage is reduced. Some ser‐
2117 vice levels are lower, notable very large data and a high TCP load are
2118 no longer supported. Very large data and high TCP loads are exceptional
2119 for the DNS. DNSSEC validation is enabled, just add trust anchors. If
2120 you do not have to worry about programs using more than 3 Mb of memory,
2121 the below example is not for you. Use the defaults to receive full ser‐
2122 vice, which on BSD-32bit tops out at 30-40 Mb after heavy usage.
2123 # example settings that reduce memory usage
2125 server:
2126 num-threads: 1
2127 outgoing-num-tcp: 1 # this limits TCP service, uses less buffers.
2128 incoming-num-tcp: 1
2129 outgoing-range: 60 # uses less memory, but less performance.
2130 msg-buffer-size: 8192 # note this limits service, 'no huge stuff'.
2131 msg-cache-size: 100k
2132 msg-cache-slabs: 1
2133 rrset-cache-size: 100k
2134 rrset-cache-slabs: 1
2135 infra-cache-numhosts: 200
2136 infra-cache-slabs: 1
2137 key-cache-size: 100k
2138 key-cache-slabs: 1
2139 neg-cache-size: 10k
2140 num-queries-per-thread: 30
2141 target-fetch-policy: "2 1 0 0 0 0"
2142 harden-large-queries: "yes"
2143 harden-short-bufsize: "yes"
2144
2146 /etc/unbound
2147 default unbound working directory.
2148 /etc/unbound
2150 default chroot(2) location.
2151 /etc/unbound/unbound.conf
2153 unbound configuration file.
2154 /var/run/unbound/unbound.pid
2156 default unbound pidfile with process ID of the running daemon.
2157 unbound.log
2159 unbound log file. default is to log to syslog(3).
2160
2162 unbound(8), unbound-checkconf(8).
2163
2165 Unbound was written by NLnet Labs. Please see CREDITS file in the dis‐
2166 tribution for further details.
2167NLnet Labs Dec 11, 2018 unbound.conf(5)