1unbound.conf(5) unbound 1.13.2 unbound.conf(5)
2
3
4
6 unbound.conf - Unbound configuration file.
7
9 unbound.conf
10
12 unbound.conf is used to configure unbound(8). The file format has at‐
13 tributes and values. Some attributes have attributes inside them. The
14 notation is: attribute: value.
15
16 Comments start with # and last to the end of line. Empty lines are ig‐
17 nored as is whitespace at the beginning of a line.
18
19 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 /etc/unbound/un‐
24 bound.conf and start the server with:
25
26 $ unbound -c /etc/unbound/unbound.conf
27
28 Most settings are the defaults. Stop the server with:
29
30 $ kill `cat /etc/unbound/unbound.pid`
31
32 Below is a minimal config file. The source distribution contains an ex‐
33 tensive example.conf file with all the options.
34
35 # 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/urandom /etc/unbound/dev/urandom
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 a value, or its containing
56 attributes in which case it is referred to as a clause. Clauses can be
57 repeated throughout the file (or included files) to group attributes
58 under the same clause.
59
60 Files can be included using the include: directive. It can appear any‐
61 where, it accepts a single file name as argument. Processing continues
62 as if the text from the included file was copied into the config file
63 at that point. If also using chroot, using full path names for the in‐
64 cluded files works, relative pathnames for the included names work if
65 the directory where the daemon is started equals its chroot/working di‐
66 rectory or is specified before the include statement with directory:
67 dir. Wildcards can be used to include multiple files, see glob(7).
68
69 For a more structural include option, the include-toplevel: directive
70 can be used. This closes whatever clause is currently active (if any)
71 and forces the use of clauses in the included files and right after
72 this directive.
73
74 Server Options
75 These options are part of the server: clause.
76
77 verbosity: <number>
78 The verbosity number, level 0 means no verbosity, only errors.
79 Level 1 gives operational information. Level 2 gives detailed
80 operational information including short information per query.
81 Level 3 gives query level information, output per query. Level
82 4 gives algorithm level information. Level 5 logs client iden‐
83 tification for cache misses. Default is level 1. The verbosity
84 can also be increased from the commandline, see unbound(8).
85
86 statistics-interval: <seconds>
87 The number of seconds between printing statistics to the log for
88 every thread. Disable with value 0 or "". Default is disabled.
89 The histogram statistics are only printed if replies were sent
90 during the statistics interval, requestlist statistics are
91 printed for every interval (but can be 0). This is because the
92 median calculation requires data to be present.
93
94 statistics-cumulative: <yes or no>
95 If enabled, statistics are cumulative since starting unbound,
96 without clearing the statistics counters after logging the sta‐
97 tistics. Default is no.
98
99 extended-statistics: <yes or no>
100 If enabled, extended statistics are printed from unbound-con‐
101 trol(8). Default is off, because keeping track of more statis‐
102 tics takes time. The counters are listed in unbound-control(8).
103
104 num-threads: <number>
105 The number of threads to create to serve clients. Use 1 for no
106 threading.
107
108 port: <port number>
109 The port number, default 53, on which the server responds to
110 queries.
111
112 interface: <ip address[@port]>
113 Interface to use to connect to the network. This interface is
114 listened to for queries from clients, and answers to clients are
115 given from it. Can be given multiple times to work on several
116 interfaces. If none are given the default is to listen to local‐
117 host. If an interface name is used instead of an ip address,
118 the list of ip addresses on that interface are used. The inter‐
119 faces are not changed on a reload (kill -HUP) but only on
120 restart. A port number can be specified with @port (without
121 spaces between interface and port number), if not specified the
122 default port (from port) is used.
123
124 ip-address: <ip address[@port]>
125 Same as interface: (for ease of compatibility with nsd.conf).
126
127 interface-automatic: <yes or no>
128 Listen on all addresses on all (current and future) interfaces,
129 detect the source interface on UDP queries and copy them to
130 replies. This is a lot like ip-transparent, but this option
131 services all interfaces whilst with ip-transparent you can se‐
132 lect which (future) interfaces unbound provides service on.
133 This feature is experimental, and needs support in your OS for
134 particular socket options. Default value is no.
135
136 outgoing-interface: <ip address or ip6 netblock>
137 Interface to use to connect to the network. This interface is
138 used to send queries to authoritative servers and receive their
139 replies. Can be given multiple times to work on several inter‐
140 faces. If none are given the default (all) is used. You can
141 specify the same interfaces in interface: and outgoing-inter‐
142 face: lines, the interfaces are then used for both purposes.
143 Outgoing queries are sent via a random outgoing interface to
144 counter spoofing.
145
146 If an IPv6 netblock is specified instead of an individual IPv6
147 address, outgoing UDP queries will use a randomised source ad‐
148 dress taken from the netblock to counter spoofing. Requires the
149 IPv6 netblock to be routed to the host running unbound, and re‐
150 quires OS support for unprivileged non-local binds (currently
151 only supported on Linux). Several netblocks may be specified
152 with multiple outgoing-interface: options, but do not specify
153 both an individual IPv6 address and an IPv6 netblock, or the
154 randomisation will be compromised. Consider combining with pre‐
155 fer-ip6: yes to increase the likelihood of IPv6 nameservers be‐
156 ing selected for queries. On Linux you need these two commands
157 to be able to use the freebind socket option to receive traffic
158 for the ip6 netblock: ip -6 addr add mynetblock/64 dev lo && ip
159 -6 route add local mynetblock/64 dev lo
160
161 outgoing-range: <number>
162 Number of ports to open. This number of file descriptors can be
163 opened per thread. Must be at least 1. Default depends on com‐
164 pile options. Larger numbers need extra resources from the oper‐
165 ating system. For performance a very large value is best, use
166 libevent to make this possible.
167
168 outgoing-port-permit: <port number or range>
169 Permit unbound to open this port or range of ports for use to
170 send queries. A larger number of permitted outgoing ports in‐
171 creases resilience against spoofing attempts. Make sure these
172 ports are not needed by other daemons. By default only ports
173 above 1024 that have not been assigned by IANA are used. Give a
174 port number or a range of the form "low-high", without spaces.
175
176 The outgoing-port-permit and outgoing-port-avoid statements are
177 processed in the line order of the config file, adding the per‐
178 mitted ports and subtracting the avoided ports from the set of
179 allowed ports. The processing starts with the non IANA allo‐
180 cated ports above 1024 in the set of allowed ports.
181
182 outgoing-port-avoid: <port number or range>
183 Do not permit unbound to open this port or range of ports for
184 use to send queries. Use this to make sure unbound does not grab
185 a port that another daemon needs. The port is avoided on all
186 outgoing interfaces, both IP4 and IP6. By default only ports
187 above 1024 that have not been assigned by IANA are used. Give a
188 port number or a range of the form "low-high", without spaces.
189
190 outgoing-num-tcp: <number>
191 Number of outgoing TCP buffers to allocate per thread. Default
192 is 10. If set to 0, or if do-tcp is "no", no TCP queries to au‐
193 thoritative servers are done. For larger installations increas‐
194 ing this value is a good idea.
195
196 incoming-num-tcp: <number>
197 Number of incoming TCP buffers to allocate per thread. Default
198 is 10. If set to 0, or if do-tcp is "no", no TCP queries from
199 clients are accepted. For larger installations increasing this
200 value is a good idea.
201
202 edns-buffer-size: <number>
203 Number of bytes size to advertise as the EDNS reassembly buffer
204 size. This is the value put into datagrams over UDP towards
205 peers. The actual buffer size is determined by msg-buffer-size
206 (both for TCP and UDP). Do not set higher than that value. De‐
207 fault is 1232 which is the DNS Flag Day 2020 recommendation.
208 Setting to 512 bypasses even the most stringent path MTU prob‐
209 lems, but is seen as extreme, since the amount of TCP fallback
210 generated is excessive (probably also for this resolver, con‐
211 sider tuning the outgoing tcp number).
212
213 max-udp-size: <number>
214 Maximum UDP response size (not applied to TCP response). 65536
215 disables the udp response size maximum, and uses the choice from
216 the client, always. Suggested values are 512 to 4096. Default
217 is 4096.
218
219 stream-wait-size: <number>
220 Number of bytes size maximum to use for waiting stream buffers.
221 Default is 4 megabytes. A plain number is in bytes, append 'k',
222 'm' or 'g' for kilobytes, megabytes or gigabytes (1024*1024
223 bytes in a megabyte). As TCP and TLS streams queue up multiple
224 results, the amount of memory used for these buffers does not
225 exceed this number, otherwise the responses are dropped. This
226 manages the total memory usage of the server (under heavy use),
227 the number of requests that can be queued up per connection is
228 also limited, with further requests waiting in TCP buffers.
229
230 msg-buffer-size: <number>
231 Number of bytes size of the message buffers. Default is 65552
232 bytes, enough for 64 Kb packets, the maximum DNS message size.
233 No message larger than this can be sent or received. Can be re‐
234 duced to use less memory, but some requests for DNS data, such
235 as for huge resource records, will result in a SERVFAIL reply to
236 the client.
237
238 msg-cache-size: <number>
239 Number of bytes size of the message cache. Default is 4
240 megabytes. A plain number is in bytes, append 'k', 'm' or 'g'
241 for kilobytes, megabytes or gigabytes (1024*1024 bytes in a
242 megabyte).
243
244 msg-cache-slabs: <number>
245 Number of slabs in the message cache. Slabs reduce lock con‐
246 tention by threads. Must be set to a power of 2. Setting
247 (close) to the number of cpus is a reasonable guess.
248
249 num-queries-per-thread: <number>
250 The number of queries that every thread will service simultane‐
251 ously. If more queries arrive that need servicing, and no
252 queries can be jostled out (see jostle-timeout), then the
253 queries are dropped. This forces the client to resend after a
254 timeout; allowing the server time to work on the existing
255 queries. Default depends on compile options, 512 or 1024.
256
257 jostle-timeout: <msec>
258 Timeout used when the server is very busy. Set to a value that
259 usually results in one roundtrip to the authority servers. If
260 too many queries arrive, then 50% of the queries are allowed to
261 run to completion, and the other 50% are replaced with the new
262 incoming query if they have already spent more than their al‐
263 lowed time. This protects against denial of service by slow
264 queries or high query rates. Default 200 milliseconds. The ef‐
265 fect is that the qps for long-lasting queries is about (num‐
266 queriesperthread / 2) / (average time for such long queries)
267 qps. The qps for short queries can be about (numqueries‐
268 perthread / 2) / (jostletimeout in whole seconds) qps per
269 thread, about (1024/2)*5 = 2560 qps by default.
270
271 delay-close: <msec>
272 Extra delay for timeouted UDP ports before they are closed, in
273 msec. Default is 0, and that disables it. This prevents very
274 delayed answer packets from the upstream (recursive) servers
275 from bouncing against closed ports and setting off all sort of
276 close-port counters, with eg. 1500 msec. When timeouts happen
277 you need extra sockets, it checks the ID and remote IP of pack‐
278 ets, and unwanted packets are added to the unwanted packet
279 counter.
280
281 udp-connect: <yes or no>
282 Perform connect for UDP sockets that mitigates ICMP side channel
283 leakage. Default is yes.
284
285 unknown-server-time-limit: <msec>
286 The wait time in msec for waiting for an unknown server to re‐
287 ply. Increase this if you are behind a slow satellite link, to
288 eg. 1128. That would then avoid re-querying every initial query
289 because it times out. Default is 376 msec.
290
291 so-rcvbuf: <number>
292 If not 0, then set the SO_RCVBUF socket option to get more buf‐
293 fer space on UDP port 53 incoming queries. So that short spikes
294 on busy servers do not drop packets (see counter in netstat
295 -su). Default is 0 (use system value). Otherwise, the number
296 of bytes to ask for, try "4m" on a busy server. The OS caps it
297 at a maximum, on linux unbound needs root permission to bypass
298 the limit, or the admin can use sysctl net.core.rmem_max. On
299 BSD change kern.ipc.maxsockbuf in /etc/sysctl.conf. On OpenBSD
300 change header and recompile kernel. On Solaris ndd -set /dev/udp
301 udp_max_buf 8388608.
302
303 so-sndbuf: <number>
304 If not 0, then set the SO_SNDBUF socket option to get more buf‐
305 fer space on UDP port 53 outgoing queries. This for very busy
306 servers handles spikes in answer traffic, otherwise 'send: re‐
307 source temporarily unavailable' can get logged, the buffer over‐
308 run is also visible by netstat -su. Default is 0 (use system
309 value). Specify the number of bytes to ask for, try "4m" on a
310 very busy server. The OS caps it at a maximum, on linux unbound
311 needs root permission to bypass the limit, or the admin can use
312 sysctl net.core.wmem_max. On BSD, Solaris changes are similar
313 to so-rcvbuf.
314
315 so-reuseport: <yes or no>
316 If yes, then open dedicated listening sockets for incoming
317 queries for each thread and try to set the SO_REUSEPORT socket
318 option on each socket. May distribute incoming queries to
319 threads more evenly. Default is yes. On Linux it is supported
320 in kernels >= 3.9. On other systems, FreeBSD, OSX it may also
321 work. You can enable it (on any platform and kernel), it then
322 attempts to open the port and passes the option if it was avail‐
323 able at compile time, if that works it is used, if it fails, it
324 continues silently (unless verbosity 3) without the option. At
325 extreme load it could be better to turn it off to distribute the
326 queries evenly, reported for Linux systems (4.4.x).
327
328 ip-transparent: <yes or no>
329 If yes, then use IP_TRANSPARENT socket option on sockets where
330 unbound is listening for incoming traffic. Default no. Allows
331 you to bind to non-local interfaces. For example for non-exis‐
332 tent IP addresses that are going to exist later on, with host
333 failover configuration. This is a lot like interface-automatic,
334 but that one services all interfaces and with this option you
335 can select which (future) interfaces unbound provides service
336 on. This option needs unbound to be started with root permis‐
337 sions on some systems. The option uses IP_BINDANY on FreeBSD
338 systems and SO_BINDANY on OpenBSD systems.
339
340 ip-freebind: <yes or no>
341 If yes, then use IP_FREEBIND socket option on sockets where un‐
342 bound is listening to incoming traffic. Default no. Allows you
343 to bind to IP addresses that are nonlocal or do not exist, like
344 when the network interface or IP address is down. Exists only
345 on Linux, where the similar ip-transparent option is also avail‐
346 able.
347
348 ip-dscp: <number>
349 The value of the Differentiated Services Codepoint (DSCP) in the
350 differentiated services field (DS) of the outgoing IP packet
351 headers. The field replaces the outdated IPv4 Type-Of-Service
352 field and the IPV6 traffic class field.
353
354 rrset-cache-size: <number>
355 Number of bytes size of the RRset cache. Default is 4 megabytes.
356 A plain number is in bytes, append 'k', 'm' or 'g' for kilo‐
357 bytes, megabytes or gigabytes (1024*1024 bytes in a megabyte).
358
359 rrset-cache-slabs: <number>
360 Number of slabs in the RRset cache. Slabs reduce lock contention
361 by threads. Must be set to a power of 2.
362
363 cache-max-ttl: <seconds>
364 Time to live maximum for RRsets and messages in the cache. De‐
365 fault is 86400 seconds (1 day). When the TTL expires, the cache
366 item has expired. Can be set lower to force the resolver to
367 query for data often, and not trust (very large) TTL values.
368 Downstream clients also see the lower TTL.
369
370 cache-min-ttl: <seconds>
371 Time to live minimum for RRsets and messages in the cache. De‐
372 fault is 0. If the minimum kicks in, the data is cached for
373 longer than the domain owner intended, and thus less queries are
374 made to look up the data. Zero makes sure the data in the cache
375 is as the domain owner intended, higher values, especially more
376 than an hour or so, can lead to trouble as the data in the cache
377 does not match up with the actual data any more.
378
379 cache-max-negative-ttl: <seconds>
380 Time to live maximum for negative responses, these have a SOA in
381 the authority section that is limited in time. Default is 3600.
382 This applies to nxdomain and nodata answers.
383
384 infra-host-ttl: <seconds>
385 Time to live for entries in the host cache. The host cache con‐
386 tains roundtrip timing, lameness and EDNS support information.
387 Default is 900.
388
389 infra-cache-slabs: <number>
390 Number of slabs in the infrastructure cache. Slabs reduce lock
391 contention by threads. Must be set to a power of 2.
392
393 infra-cache-numhosts: <number>
394 Number of hosts for which information is cached. Default is
395 10000.
396
397 infra-cache-min-rtt: <msec>
398 Lower limit for dynamic retransmit timeout calculation in infra‐
399 structure cache. Default is 50 milliseconds. Increase this value
400 if using forwarders needing more time to do recursive name reso‐
401 lution.
402
403 infra-keep-probing: <yes or no>
404 If enabled the server keeps probing hosts that are down, in the
405 one probe at a time regime. Default is no. Hosts that are
406 down, eg. they did not respond during the one probe at a time
407 period, are marked as down and it may take infra-host-ttl time
408 to get probed again.
409
410 define-tag: <"list of tags">
411 Define the tags that can be used with local-zone and access-con‐
412 trol. Enclose the list between quotes ("") and put spaces be‐
413 tween tags.
414
415 do-ip4: <yes or no>
416 Enable or disable whether ip4 queries are answered or issued.
417 Default is yes.
418
419 do-ip6: <yes or no>
420 Enable or disable whether ip6 queries are answered or issued.
421 Default is yes. If disabled, queries are not answered on IPv6,
422 and queries are not sent on IPv6 to the internet nameservers.
423 With this option you can disable the ipv6 transport for sending
424 DNS traffic, it does not impact the contents of the DNS traffic,
425 which may have ip4 and ip6 addresses in it.
426
427 prefer-ip4: <yes or no>
428 If enabled, prefer IPv4 transport for sending DNS queries to in‐
429 ternet nameservers. Default is no. Useful if the IPv6 netblock
430 the server has, the entire /64 of that is not owned by one oper‐
431 ator and the reputation of the netblock /64 is an issue, using
432 IPv4 then uses the IPv4 filters that the upstream servers have.
433
434 prefer-ip6: <yes or no>
435 If enabled, prefer IPv6 transport for sending DNS queries to in‐
436 ternet nameservers. Default is no.
437
438 do-udp: <yes or no>
439 Enable or disable whether UDP queries are answered or issued.
440 Default is yes.
441
442 do-tcp: <yes or no>
443 Enable or disable whether TCP queries are answered or issued.
444 Default is yes.
445
446 tcp-mss: <number>
447 Maximum segment size (MSS) of TCP socket on which the server re‐
448 sponds to queries. Value lower than common MSS on Ethernet (1220
449 for example) will address path MTU problem. Note that not all
450 platform supports socket option to set MSS (TCP_MAXSEG). De‐
451 fault is system default MSS determined by interface MTU and ne‐
452 gotiation between server and client.
453
454 outgoing-tcp-mss: <number>
455 Maximum segment size (MSS) of TCP socket for outgoing queries
456 (from Unbound to other servers). Value lower than common MSS on
457 Ethernet (1220 for example) will address path MTU problem. Note
458 that not all platform supports socket option to set MSS
459 (TCP_MAXSEG). Default is system default MSS determined by in‐
460 terface MTU and negotiation between Unbound and other servers.
461
462 tcp-idle-timeout: <msec>
463 The period Unbound will wait for a query on a TCP connection.
464 If this timeout expires Unbound closes the connection. This op‐
465 tion defaults to 30000 milliseconds. When the number of free
466 incoming TCP buffers falls below 50% of the total number config‐
467 ured, the option value used is progressively reduced, first to
468 1% of the configured value, then to 0.2% of the configured value
469 if the number of free buffers falls below 35% of the total num‐
470 ber configured, and finally to 0 if the number of free buffers
471 falls below 20% of the total number configured. A minimum time‐
472 out of 200 milliseconds is observed regardless of the option
473 value used.
474
475 tcp-reuse-timeout: <msec>
476 The period Unbound will keep TCP persistent connections open to
477 authority servers. This option defaults to 60000 milliseconds.
478
479 max-reuse-tcp-queries: <number>
480 The maximum number of queries that can be sent on a persistent
481 TCP connection. This option defaults to 200 queries.
482
483 tcp-auth-query-timeout: <number>
484 Timeout in milliseconds for TCP queries to auth servers. This
485 option defaults to 3000 milliseconds.
486
487 edns-tcp-keepalive: <yes or no>
488 Enable or disable EDNS TCP Keepalive. Default is no.
489
490 edns-tcp-keepalive-timeout: <msec>
491 The period Unbound will wait for a query on a TCP connection
492 when EDNS TCP Keepalive is active. If this timeout expires Un‐
493 bound closes the connection. If the client supports the EDNS TCP
494 Keepalive option, Unbound sends the timeout value to the client
495 to encourage it to close the connection before the server times
496 out. This option defaults to 120000 milliseconds. When the
497 number of free incoming TCP buffers falls below 50% of the total
498 number configured, the advertised timeout is progressively re‐
499 duced to 1% of the configured value, then to 0.2% of the config‐
500 ured value if the number of free buffers falls below 35% of the
501 total number configured, and finally to 0 if the number of free
502 buffers falls below 20% of the total number configured. A mini‐
503 mum actual timeout of 200 milliseconds is observed regardless of
504 the advertised timeout.
505
506 tcp-upstream: <yes or no>
507 Enable or disable whether the upstream queries use TCP only for
508 transport. Default is no. Useful in tunneling scenarios.
509
510 udp-upstream-without-downstream: <yes or no>
511 Enable udp upstream even if do-udp is no. Default is no, and
512 this does not change anything. Useful for TLS service
513 providers, that want no udp downstream but use udp to fetch data
514 upstream.
515
516 tls-upstream: <yes or no>
517 Enabled or disable whether the upstream queries use TLS only for
518 transport. Default is no. Useful in tunneling scenarios. The
519 TLS contains plain DNS in TCP wireformat. The other server must
520 support this (see tls-service-key). If you enable this, also
521 configure a tls-cert-bundle or use tls-win-cert to load CA
522 certs, otherwise the connections cannot be authenticated. This
523 option enables TLS for all of them, but if you do not set this
524 you can configure TLS specifically for some forward zones with
525 forward-tls-upstream. And also with stub-tls-upstream.
526
527 ssl-upstream: <yes or no>
528 Alternate syntax for tls-upstream. If both are present in the
529 config file the last is used.
530
531 tls-service-key: <file>
532 If enabled, the server provides DNS-over-TLS or DNS-over-HTTPS
533 service on the TCP ports marked implicitly or explicitly for
534 these services with tls-port or https-port. The file must con‐
535 tain the private key for the TLS session, the public certificate
536 is in the tls-service-pem file and it must also be specified if
537 tls-service-key is specified. The default is "", turned off.
538 Enabling or disabling this service requires a restart (a reload
539 is not enough), because the key is read while root permissions
540 are held and before chroot (if any). The ports enabled implic‐
541 itly or explicitly via tls-port: and https-port: do not provide
542 normal DNS TCP service. Unbound needs to be compiled with lib‐
543 nghttp2 in order to provide DNS-over-HTTPS.
544
545 ssl-service-key: <file>
546 Alternate syntax for tls-service-key.
547
548 tls-service-pem: <file>
549 The public key certificate pem file for the tls service. De‐
550 fault is "", turned off.
551
552 ssl-service-pem: <file>
553 Alternate syntax for tls-service-pem.
554
555 tls-port: <number>
556 The port number on which to provide TCP TLS service, default
557 853, only interfaces configured with that port number as @number
558 get the TLS service.
559
560 ssl-port: <number>
561 Alternate syntax for tls-port.
562
563 tls-cert-bundle: <file>
564 If null or "", no file is used. Set it to the certificate bun‐
565 dle file, for example "/etc/pki/tls/certs/ca-bundle.crt". These
566 certificates are used for authenticating connections made to
567 outside peers. For example auth-zone urls, and also DNS over
568 TLS connections. It is read at start up before permission drop
569 and chroot.
570
571 ssl-cert-bundle: <file>
572 Alternate syntax for tls-cert-bundle.
573
574 tls-win-cert: <yes or no>
575 Add the system certificates to the cert bundle certificates for
576 authentication. If no cert bundle, it uses only these certifi‐
577 cates. Default is no. On windows this option uses the certifi‐
578 cates from the cert store. Use the tls-cert-bundle option on
579 other systems.
580
581 tls-additional-port: <portnr>
582 List portnumbers as tls-additional-port, and when interfaces are
583 defined, eg. with the @port suffix, as this port number, they
584 provide dns over TLS service. Can list multiple, each on a new
585 statement.
586
587 tls-session-ticket-keys: <file>
588 If not "", lists files with 80 bytes of random contents that are
589 used to perform TLS session resumption for clients using the un‐
590 bound server. These files contain the secret key for the TLS
591 session tickets. First key use to encrypt and decrypt TLS ses‐
592 sion tickets. Other keys use to decrypt only. With this you
593 can roll over to new keys, by generating a new first file and
594 allowing decrypt of the old file by listing it after the first
595 file for some time, after the wait clients are not using the old
596 key any more and the old key can be removed. One way to create
597 the file is dd if=/dev/random bs=1 count=80 of=ticket.dat The
598 first 16 bytes should be different from the old one if you cre‐
599 ate a second key, that is the name used to identify the key.
600 Then there is 32 bytes random data for an AES key and then 32
601 bytes random data for the HMAC key.
602
603 tls-ciphers: <string with cipher list>
604 Set the list of ciphers to allow when serving TLS. Use "" for
605 defaults, and that is the default.
606
607 tls-ciphersuites: <string with ciphersuites list>
608 Set the list of ciphersuites to allow when serving TLS. This is
609 for newer TLS 1.3 connections. Use "" for defaults, and that is
610 the default.
611
612 pad-responses: <yes or no>
613 If enabled, TLS serviced queries that contained an EDNS Padding
614 option will cause responses padded to the closest multiple of
615 the size specified in pad-responses-block-size. Default is yes.
616
617 pad-responses-block-size: <number>
618 The block size with which to pad responses serviced over TLS.
619 Only responses to padded queries will be padded. Default is
620 468.
621
622 pad-queries: <yes or no>
623 If enabled, all queries sent over TLS upstreams will be padded
624 to the closest multiple of the size specified in
625 pad-queries-block-size. Default is yes.
626
627 pad-queries-block-size: <number>
628 The block size with which to pad queries sent over TLS up‐
629 streams. Default is 128.
630
631 tls-use-sni: <yes or no>
632 Enable or disable sending the SNI extension on TLS connections.
633 Default is yes. Changing the value requires a reload.
634
635 https-port: <number>
636 The port number on which to provide DNS-over-HTTPS service, de‐
637 fault 443, only interfaces configured with that port number as
638 @number get the HTTPS service.
639
640 http-endpoint: <endpoint string>
641 The HTTP endpoint to provide DNS-over-HTTPS service on. Default
642 "/dns-query".
643
644 http-max-streams: <number of streams>
645 Number used in the SETTINGS_MAX_CONCURRENT_STREAMS parameter in
646 the HTTP/2 SETTINGS frame for DNS-over-HTTPS connections. De‐
647 fault 100.
648
649 http-query-buffer-size: <size in bytes>
650 Maximum number of bytes used for all HTTP/2 query buffers com‐
651 bined. These buffers contain (partial) DNS queries waiting for
652 request stream completion. An RST_STREAM frame will be send to
653 streams exceeding this limit. Default is 4 megabytes. A plain
654 number is in bytes, append 'k', 'm' or 'g' for kilobytes,
655 megabytes or gigabytes (1024*1024 bytes in a megabyte).
656
657 http-response-buffer-size: <size in bytes>
658 Maximum number of bytes used for all HTTP/2 response buffers
659 combined. These buffers contain DNS responses waiting to be
660 written back to the clients. An RST_STREAM frame will be send
661 to streams exceeding this limit. Default is 4 megabytes. A plain
662 number is in bytes, append 'k', 'm' or 'g' for kilobytes,
663 megabytes or gigabytes (1024*1024 bytes in a megabyte).
664
665 http-nodelay: <yes or no>
666 Set TCP_NODELAY socket option on sockets used to provide DNS-
667 over-HTTPS service. Ignored if the option is not available. De‐
668 fault is yes.
669
670 http-notls-downstream: <yes or no>
671 Disable use of TLS for the downstream DNS-over-HTTP connections.
672 Useful for local back end servers. Default is no.
673
674 use-systemd: <yes or no>
675 Enable or disable systemd socket activation. Default is no.
676
677 do-daemonize: <yes or no>
678 Enable or disable whether the unbound server forks into the
679 background as a daemon. Set the value to no when unbound runs
680 as systemd service. Default is yes.
681
682 tcp-connection-limit: <IP netblock> <limit>
683 Allow up to limit simultaneous TCP connections from the given
684 netblock. When at the limit, further connections are accepted
685 but closed immediately. This option is experimental at this
686 time.
687
688 access-control: <IP netblock> <action>
689 The netblock is given as an IP4 or IP6 address with /size ap‐
690 pended for a classless network block. The action can be deny,
691 refuse, allow, allow_setrd, allow_snoop, deny_non_local or
692 refuse_non_local. The most specific netblock match is used, if
693 none match deny is used. The order of the access-control state‐
694 ments therefore does not matter.
695
696 The action deny stops queries from hosts from that netblock.
697
698 The action refuse stops queries too, but sends a DNS rcode RE‐
699 FUSED error message back.
700
701 The action allow gives access to clients from that netblock. It
702 gives only access for recursion clients (which is what almost
703 all clients need). Nonrecursive queries are refused.
704
705 The allow action does allow nonrecursive queries to access the
706 local-data that is configured. The reason is that this does not
707 involve the unbound server recursive lookup algorithm, and
708 static data is served in the reply. This supports normal opera‐
709 tions where nonrecursive queries are made for the authoritative
710 data. For nonrecursive queries any replies from the dynamic
711 cache are refused.
712
713 The allow_setrd action ignores the recursion desired (RD) bit
714 and treats all requests as if the recursion desired bit is set.
715 Note that this behavior violates RFC 1034 which states that a
716 name server should never perform recursive service unless asked
717 via the RD bit since this interferes with trouble shooting of
718 name servers and their databases. This prohibited behavior may
719 be useful if another DNS server must forward requests for spe‐
720 cific zones to a resolver DNS server, but only supports stub do‐
721 mains and sends queries to the resolver DNS server with the RD
722 bit cleared.
723
724 The action allow_snoop gives nonrecursive access too. This give
725 both recursive and non recursive access. The name allow_snoop
726 refers to cache snooping, a technique to use nonrecursive
727 queries to examine the cache contents (for malicious acts).
728 However, nonrecursive queries can also be a valuable debugging
729 tool (when you want to examine the cache contents). In that case
730 use allow_snoop for your administration host.
731
732 By default only localhost is allowed, the rest is refused. The
733 default is refused, because that is protocol-friendly. The DNS
734 protocol is not designed to handle dropped packets due to pol‐
735 icy, and dropping may result in (possibly excessive) retried
736 queries.
737
738 The deny_non_local and refuse_non_local settings are for hosts
739 that are only allowed to query for the authoritative local-data,
740 they are not allowed full recursion but only the static data.
741 With deny_non_local, messages that are disallowed are dropped,
742 with refuse_non_local they receive error code REFUSED.
743
744 access-control-tag: <IP netblock> <"list of tags">
745 Assign tags to access-control elements. Clients using this ac‐
746 cess control element use localzones that are tagged with one of
747 these tags. Tags must be defined in define-tags. Enclose list
748 of tags in quotes ("") and put spaces between tags. If ac‐
749 cess-control-tag is configured for a netblock that does not have
750 an access-control, an access-control element with action allow
751 is configured for this netblock.
752
753 access-control-tag-action: <IP netblock> <tag> <action>
754 Set action for particular tag for given access control element.
755 If you have multiple tag values, the tag used to lookup the ac‐
756 tion is the first tag match between access-control-tag and lo‐
757 cal-zone-tag where "first" comes from the order of the define-
758 tag values.
759
760 access-control-tag-data: <IP netblock> <tag> <"resource record string">
761 Set redirect data for particular tag for given access control
762 element.
763
764 access-control-view: <IP netblock> <view name>
765 Set view for given access control element.
766
767 chroot: <directory>
768 If chroot is enabled, you should pass the configfile (from the
769 commandline) as a full path from the original root. After the
770 chroot has been performed the now defunct portion of the config
771 file path is removed to be able to reread the config after a
772 reload.
773
774 All other file paths (working dir, logfile, roothints, and key
775 files) can be specified in several ways: as an absolute path
776 relative to the new root, as a relative path to the working di‐
777 rectory, or as an absolute path relative to the original root.
778 In the last case the path is adjusted to remove the unused por‐
779 tion.
780
781 The pidfile can be either a relative path to the working direc‐
782 tory, or an absolute path relative to the original root. It is
783 written just prior to chroot and dropping permissions. This al‐
784 lows the pidfile to be /var/run/unbound.pid and the chroot to be
785 /var/unbound, for example. Note that Unbound is not able to re‐
786 move the pidfile after termination when it is located outside of
787 the chroot directory.
788
789 Additionally, unbound may need to access /dev/urandom (for en‐
790 tropy) from inside the chroot.
791
792 If given a chroot is done to the given directory. By default ch‐
793 root is enabled and the default is "/etc/unbound". If you give
794 "" no chroot is performed.
795
796 username: <name>
797 If given, after binding the port the user privileges are
798 dropped. Default is "unbound". If you give username: "" no user
799 change is performed.
800
801 If this user is not capable of binding the port, reloads (by
802 signal HUP) will still retain the opened ports. If you change
803 the port number in the config file, and that new port number re‐
804 quires privileges, then a reload will fail; a restart is needed.
805
806 directory: <directory>
807 Sets the working directory for the program. Default is "/etc/un‐
808 bound". On Windows the string "%EXECUTABLE%" tries to change to
809 the directory that unbound.exe resides in. If you give a
810 server: directory: dir before include: file statements then
811 those includes can be relative to the working directory.
812
813 logfile: <filename>
814 If "" is given, logging goes to stderr, or nowhere once daemo‐
815 nized. The logfile is appended to, in the following format:
816 [seconds since 1970] unbound[pid:tid]: type: message.
817 If this option is given, the use-syslog is option is set to
818 "no". The logfile is reopened (for append) when the config file
819 is reread, on SIGHUP.
820
821 use-syslog: <yes or no>
822 Sets unbound to send log messages to the syslogd, using sys‐
823 log(3). The log facility LOG_DAEMON is used, with identity "un‐
824 bound". The logfile setting is overridden when use-syslog is
825 turned on. The default is to log to syslog.
826
827 log-identity: <string>
828 If "" is given (default), then the name of the executable, usu‐
829 ally "unbound" is used to report to the log. Enter a string to
830 override it with that, which is useful on systems that run more
831 than one instance of unbound, with different configurations, so
832 that the logs can be easily distinguished against.
833
834 log-time-ascii: <yes or no>
835 Sets logfile lines to use a timestamp in UTC ascii. Default is
836 no, which prints the seconds since 1970 in brackets. No effect
837 if using syslog, in that case syslog formats the timestamp
838 printed into the log files.
839
840 log-queries: <yes or no>
841 Prints one line per query to the log, with the log timestamp and
842 IP address, name, type and class. Default is no. Note that it
843 takes time to print these lines which makes the server (signifi‐
844 cantly) slower. Odd (nonprintable) characters in names are
845 printed as '?'.
846
847 log-replies: <yes or no>
848 Prints one line per reply to the log, with the log timestamp and
849 IP address, name, type, class, return code, time to resolve,
850 from cache and response size. Default is no. Note that it
851 takes time to print these lines which makes the server (signifi‐
852 cantly) slower. Odd (nonprintable) characters in names are
853 printed as '?'.
854
855 log-tag-queryreply: <yes or no>
856 Prints the word 'query' and 'reply' with log-queries and
857 log-replies. This makes filtering logs easier. The default is
858 off (for backwards compatibility).
859
860 log-local-actions: <yes or no>
861 Print log lines to inform about local zone actions. These lines
862 are like the local-zone type inform prints out, but they are
863 also printed for the other types of local zones.
864
865 log-servfail: <yes or no>
866 Print log lines that say why queries return SERVFAIL to clients.
867 This is separate from the verbosity debug logs, much smaller,
868 and printed at the error level, not the info level of debug info
869 from verbosity.
870
871 pidfile: <filename>
872 The process id is written to the file. Default is "/run/un‐
873 bound/unbound.pid". So,
874 kill -HUP `cat /run/unbound/unbound.pid`
875 triggers a reload,
876 kill -TERM `cat /run/unbound/unbound.pid`
877 gracefully terminates.
878
879 root-hints: <filename>
880 Read the root hints from this file. Default is nothing, using
881 builtin hints for the IN class. The file has the format of zone
882 files, with root nameserver names and addresses only. The de‐
883 fault may become outdated, when servers change, therefore it is
884 good practice to use a root-hints file.
885
886 hide-identity: <yes or no>
887 If enabled id.server and hostname.bind queries are refused.
888
889 identity: <string>
890 Set the identity to report. If set to "", the default, then the
891 hostname of the server is returned.
892
893 hide-version: <yes or no>
894 If enabled version.server and version.bind queries are refused.
895
896 version: <string>
897 Set the version to report. If set to "", the default, then the
898 package version is returned.
899
900 hide-http-user-agent: <yes or no>
901 If enabled the HTTP header User-Agent is not set. Use with cau‐
902 tion as some webserver configurations may reject HTTP requests
903 lacking this header. If needed, it is better to explicitly set
904 the http-user-agent below.
905
906 http-user-agent: <string>
907 Set the HTTP User-Agent header for outgoing HTTP requests. If
908 set to "", the default, then the package name and version are
909 used.
910
911 nsid: <string>
912 Add the specified nsid to the EDNS section of the answer when
913 queried with an NSID EDNS enabled packet. As a sequence of hex
914 characters or with ascii_ prefix and then an ascii string.
915
916 hide-trustanchor: <yes or no>
917 If enabled trustanchor.unbound queries are refused.
918
919 target-fetch-policy: <"list of numbers">
920 Set the target fetch policy used by unbound to determine if it
921 should fetch nameserver target addresses opportunistically. The
922 policy is described per dependency depth.
923
924 The number of values determines the maximum dependency depth
925 that unbound will pursue in answering a query. A value of -1
926 means to fetch all targets opportunistically for that dependency
927 depth. A value of 0 means to fetch on demand only. A positive
928 value fetches that many targets opportunistically.
929
930 Enclose the list between quotes ("") and put spaces between num‐
931 bers. The default is "3 2 1 0 0". Setting all zeroes, "0 0 0 0
932 0" gives behaviour closer to that of BIND 9, while setting "-1
933 -1 -1 -1 -1" gives behaviour rumoured to be closer to that of
934 BIND 8.
935
936 harden-short-bufsize: <yes or no>
937 Very small EDNS buffer sizes from queries are ignored. Default
938 is on, as described in the standard.
939
940 harden-large-queries: <yes or no>
941 Very large queries are ignored. Default is off, since it is le‐
942 gal protocol wise to send these, and could be necessary for op‐
943 eration if TSIG or EDNS payload is very large.
944
945 harden-glue: <yes or no>
946 Will trust glue only if it is within the servers authority. De‐
947 fault is yes.
948
949 harden-dnssec-stripped: <yes or no>
950 Require DNSSEC data for trust-anchored zones, if such data is
951 absent, the zone becomes bogus. If turned off, and no DNSSEC
952 data is received (or the DNSKEY data fails to validate), then
953 the zone is made insecure, this behaves like there is no trust
954 anchor. You could turn this off if you are sometimes behind an
955 intrusive firewall (of some sort) that removes DNSSEC data from
956 packets, or a zone changes from signed to unsigned to badly
957 signed often. If turned off you run the risk of a downgrade at‐
958 tack that disables security for a zone. Default is yes.
959
960 harden-below-nxdomain: <yes or no>
961 From RFC 8020 (with title "NXDOMAIN: There Really Is Nothing Un‐
962 derneath"), returns nxdomain to queries for a name below another
963 name that is already known to be nxdomain. DNSSEC mandates no‐
964 error for empty nonterminals, hence this is possible. Very old
965 software might return nxdomain for empty nonterminals (that usu‐
966 ally happen for reverse IP address lookups), and thus may be in‐
967 compatible with this. To try to avoid this only DNSSEC-secure
968 nxdomains are used, because the old software does not have
969 DNSSEC. Default is yes. The nxdomain must be secure, this
970 means nsec3 with optout is insufficient.
971
972 harden-referral-path: <yes or no>
973 Harden the referral path by performing additional queries for
974 infrastructure data. Validates the replies if trust anchors are
975 configured and the zones are signed. This enforces DNSSEC vali‐
976 dation on nameserver NS sets and the nameserver addresses that
977 are encountered on the referral path to the answer. Default no,
978 because it burdens the authority servers, and it is not RFC
979 standard, and could lead to performance problems because of the
980 extra query load that is generated. Experimental option. If
981 you enable it consider adding more numbers after the tar‐
982 get-fetch-policy to increase the max depth that is checked to.
983
984 harden-algo-downgrade: <yes or no>
985 Harden against algorithm downgrade when multiple algorithms are
986 advertised in the DS record. If no, allows the weakest algo‐
987 rithm to validate the zone. Default is no. Zone signers must
988 produce zones that allow this feature to work, but sometimes
989 they do not, and turning this option off avoids that validation
990 failure.
991
992 use-caps-for-id: <yes or no>
993 Use 0x20-encoded random bits in the query to foil spoof at‐
994 tempts. This perturbs the lowercase and uppercase of query
995 names sent to authority servers and checks if the reply still
996 has the correct casing. Disabled by default. This feature is
997 an experimental implementation of draft dns-0x20.
998
999 caps-exempt: <domain>
1000 Exempt the domain so that it does not receive caps-for-id per‐
1001 turbed queries. For domains that do not support 0x20 and also
1002 fail with fallback because they keep sending different answers,
1003 like some load balancers. Can be given multiple times, for dif‐
1004 ferent domains.
1005
1006 caps-whitelist: <yes or no>
1007 Alternate syntax for caps-exempt.
1008
1009 qname-minimisation: <yes or no>
1010 Send minimum amount of information to upstream servers to en‐
1011 hance privacy. Only send minimum required labels of the QNAME
1012 and set QTYPE to A when possible. Best effort approach; full
1013 QNAME and original QTYPE will be sent when upstream replies with
1014 a RCODE other than NOERROR, except when receiving NXDOMAIN from
1015 a DNSSEC signed zone. Default is yes.
1016
1017 qname-minimisation-strict: <yes or no>
1018 QNAME minimisation in strict mode. Do not fall-back to sending
1019 full QNAME to potentially broken nameservers. A lot of domains
1020 will not be resolvable when this option in enabled. Only use if
1021 you know what you are doing. This option only has effect when
1022 qname-minimisation is enabled. Default is no.
1023
1024 aggressive-nsec: <yes or no>
1025 Aggressive NSEC uses the DNSSEC NSEC chain to synthesize NXDO‐
1026 MAIN and other denials, using information from previous NXDO‐
1027 MAINs answers. Default is no. It helps to reduce the query
1028 rate towards targets that get a very high nonexistent name
1029 lookup rate.
1030
1031 private-address: <IP address or subnet>
1032 Give IPv4 of IPv6 addresses or classless subnets. These are ad‐
1033 dresses on your private network, and are not allowed to be re‐
1034 turned for public internet names. Any occurrence of such ad‐
1035 dresses are removed from DNS answers. Additionally, the DNSSEC
1036 validator may mark the answers bogus. This protects against
1037 so-called DNS Rebinding, where a user browser is turned into a
1038 network proxy, allowing remote access through the browser to
1039 other parts of your private network. Some names can be allowed
1040 to contain your private addresses, by default all the local-data
1041 that you configured is allowed to, and you can specify addi‐
1042 tional names using private-domain. No private addresses are en‐
1043 abled by default. We consider to enable this for the RFC1918
1044 private IP address space by default in later releases. That
1045 would enable private addresses for 10.0.0.0/8 172.16.0.0/12
1046 192.168.0.0/16 169.254.0.0/16 fd00::/8 and fe80::/10, since the
1047 RFC standards say these addresses should not be visible on the
1048 public internet. Turning on 127.0.0.0/8 would hinder many spam‐
1049 blocklists as they use that. Adding ::ffff:0:0/96 stops
1050 IPv4-mapped IPv6 addresses from bypassing the filter.
1051
1052 private-domain: <domain name>
1053 Allow this domain, and all its subdomains to contain private ad‐
1054 dresses. Give multiple times to allow multiple domain names to
1055 contain private addresses. Default is none.
1056
1057 unwanted-reply-threshold: <number>
1058 If set, a total number of unwanted replies is kept track of in
1059 every thread. When it reaches the threshold, a defensive action
1060 is taken and a warning is printed to the log. The defensive ac‐
1061 tion is to clear the rrset and message caches, hopefully flush‐
1062 ing away any poison. A value of 10 million is suggested. De‐
1063 fault is 0 (turned off).
1064
1065 do-not-query-address: <IP address>
1066 Do not query the given IP address. Can be IP4 or IP6. Append
1067 /num to indicate a classless delegation netblock, for example
1068 like 10.2.3.4/24 or 2001::11/64.
1069
1070 do-not-query-localhost: <yes or no>
1071 If yes, localhost is added to the do-not-query-address entries,
1072 both IP6 ::1 and IP4 127.0.0.1/8. If no, then localhost can be
1073 used to send queries to. Default is yes.
1074
1075 prefetch: <yes or no>
1076 If yes, message cache elements are prefetched before they expire
1077 to keep the cache up to date. Default is no. Turning it on
1078 gives about 10 percent more traffic and load on the machine, but
1079 popular items do not expire from the cache.
1080
1081 prefetch-key: <yes or no>
1082 If yes, fetch the DNSKEYs earlier in the validation process,
1083 when a DS record is encountered. This lowers the latency of re‐
1084 quests. It does use a little more CPU. Also if the cache is
1085 set to 0, it is no use. Default is no.
1086
1087 deny-any: <yes or no>
1088 If yes, deny queries of type ANY with an empty response. De‐
1089 fault is no. If disabled, unbound responds with a short list of
1090 resource records if some can be found in the cache and makes the
1091 upstream type ANY query if there are none.
1092
1093 rrset-roundrobin: <yes or no>
1094 If yes, Unbound rotates RRSet order in response (the random num‐
1095 ber is taken from the query ID, for speed and thread safety).
1096 Default is yes.
1097
1098 minimal-responses: <yes or no>
1099 If yes, Unbound does not insert authority/additional sections
1100 into response messages when those sections are not required.
1101 This reduces response size significantly, and may avoid TCP
1102 fallback for some responses. This may cause a slight speedup.
1103 The default is yes, even though the DNS protocol RFCs mandate
1104 these sections, and the additional content could be of use and
1105 save roundtrips for clients. Because they are not used, and the
1106 saved roundtrips are easier saved with prefetch, whilst this is
1107 faster.
1108
1109 disable-dnssec-lame-check: <yes or no>
1110 If true, disables the DNSSEC lameness check in the iterator.
1111 This check sees if RRSIGs are present in the answer, when dnssec
1112 is expected, and retries another authority if RRSIGs are unex‐
1113 pectedly missing. The validator will insist in RRSIGs for
1114 DNSSEC signed domains regardless of this setting, if a trust an‐
1115 chor is loaded.
1116
1117 module-config: <"module names">
1118 Module configuration, a list of module names separated by spa‐
1119 ces, surround the string with quotes (""). The modules can be
1120 respip, validator, or iterator (and possibly more, see below).
1121 Setting this to just "iterator" will result in a non-validating
1122 server. Setting this to "validator iterator" will turn on
1123 DNSSEC validation. The ordering of the modules is significant,
1124 the order decides the order of processing. You must also set
1125 trust-anchors for validation to be useful. Adding respip to the
1126 front will cause RPZ processing to be done on all queries. The
1127 default is "validator iterator".
1128
1129 When the server is built with EDNS client subnet support the de‐
1130 fault is "subnetcache validator iterator". Most modules that
1131 need to be listed here have to be listed at the beginning of the
1132 line. The subnetcachedb module has to be listed just before the
1133 iterator. The python module can be listed in different places,
1134 it then processes the output of the module it is just before.
1135 The dynlib module can be listed pretty much anywhere, it is only
1136 a very thin wrapper that allows dynamic libraries to run in its
1137 place.
1138
1139 trust-anchor-file: <filename>
1140 File with trusted keys for validation. Both DS and DNSKEY en‐
1141 tries can appear in the file. The format of the file is the
1142 standard DNS Zone file format. Default is "", or no trust an‐
1143 chor file.
1144
1145 auto-trust-anchor-file: <filename>
1146 File with trust anchor for one zone, which is tracked with
1147 RFC5011 probes. The probes are run several times per month,
1148 thus the machine must be online frequently. The initial file
1149 can be one with contents as described in trust-anchor-file. The
1150 file is written to when the anchor is updated, so the unbound
1151 user must have write permission. Write permission to the file,
1152 but also to the directory it is in (to create a temporary file,
1153 which is necessary to deal with filesystem full events), it must
1154 also be inside the chroot (if that is used).
1155
1156 trust-anchor: <"Resource Record">
1157 A DS or DNSKEY RR for a key to use for validation. Multiple en‐
1158 tries can be given to specify multiple trusted keys, in addition
1159 to the trust-anchor-files. The resource record is entered in
1160 the same format as 'dig' or 'drill' prints them, the same format
1161 as in the zone file. Has to be on a single line, with "" around
1162 it. A TTL can be specified for ease of cut and paste, but is ig‐
1163 nored. A class can be specified, but class IN is default.
1164
1165 trusted-keys-file: <filename>
1166 File with trusted keys for validation. Specify more than one
1167 file with several entries, one file per entry. Like trust-an‐
1168 chor-file but has a different file format. Format is BIND-9
1169 style format, the trusted-keys { name flag proto algo "key"; };
1170 clauses are read. It is possible to use wildcards with this
1171 statement, the wildcard is expanded on start and on reload.
1172
1173 trust-anchor-signaling: <yes or no>
1174 Send RFC8145 key tag query after trust anchor priming. Default
1175 is yes.
1176
1177 root-key-sentinel: <yes or no>
1178 Root key trust anchor sentinel. Default is yes.
1179
1180 domain-insecure: <domain name>
1181 Sets domain name to be insecure, DNSSEC chain of trust is ig‐
1182 nored towards the domain name. So a trust anchor above the do‐
1183 main name can not make the domain secure with a DS record, such
1184 a DS record is then ignored. Can be given multiple times to
1185 specify multiple domains that are treated as if unsigned. If
1186 you set trust anchors for the domain they override this setting
1187 (and the domain is secured).
1188
1189 This can be useful if you want to make sure a trust anchor for
1190 external lookups does not affect an (unsigned) internal domain.
1191 A DS record externally can create validation failures for that
1192 internal domain.
1193
1194 val-override-date: <rrsig-style date spec>
1195 Default is "" or "0", which disables this debugging feature. If
1196 enabled by giving a RRSIG style date, that date is used for ver‐
1197 ifying RRSIG inception and expiration dates, instead of the cur‐
1198 rent date. Do not set this unless you are debugging signature
1199 inception and expiration. The value -1 ignores the date alto‐
1200 gether, useful for some special applications.
1201
1202 val-sig-skew-min: <seconds>
1203 Minimum number of seconds of clock skew to apply to validated
1204 signatures. A value of 10% of the signature lifetime (expira‐
1205 tion - inception) is used, capped by this setting. Default is
1206 3600 (1 hour) which allows for daylight savings differences.
1207 Lower this value for more strict checking of short lived signa‐
1208 tures.
1209
1210 val-sig-skew-max: <seconds>
1211 Maximum number of seconds of clock skew to apply to validated
1212 signatures. A value of 10% of the signature lifetime (expira‐
1213 tion - inception) is used, capped by this setting. Default is
1214 86400 (24 hours) which allows for timezone setting problems in
1215 stable domains. Setting both min and max very low disables the
1216 clock skew allowances. Setting both min and max very high makes
1217 the validator check the signature timestamps less strictly.
1218
1219 val-max-restart: <number>
1220 The maximum number the validator should restart validation with
1221 another authority in case of failed validation. Default is 5.
1222
1223 val-bogus-ttl: <number>
1224 The time to live for bogus data. This is data that has failed
1225 validation; due to invalid signatures or other checks. The TTL
1226 from that data cannot be trusted, and this value is used in‐
1227 stead. The value is in seconds, default 60. The time interval
1228 prevents repeated revalidation of bogus data.
1229
1230 val-clean-additional: <yes or no>
1231 Instruct the validator to remove data from the additional sec‐
1232 tion of secure messages that are not signed properly. Messages
1233 that are insecure, bogus, indeterminate or unchecked are not af‐
1234 fected. Default is yes. Use this setting to protect the users
1235 that rely on this validator for authentication from potentially
1236 bad data in the additional section.
1237
1238 val-log-level: <number>
1239 Have the validator print validation failures to the log. Re‐
1240 gardless of the verbosity setting. Default is 0, off. At 1,
1241 for every user query that fails a line is printed to the logs.
1242 This way you can monitor what happens with validation. Use a
1243 diagnosis tool, such as dig or drill, to find out why validation
1244 is failing for these queries. At 2, not only the query that
1245 failed is printed but also the reason why unbound thought it was
1246 wrong and which server sent the faulty data.
1247
1248 val-permissive-mode: <yes or no>
1249 Instruct the validator to mark bogus messages as indeterminate.
1250 The security checks are performed, but if the result is bogus
1251 (failed security), the reply is not withheld from the client
1252 with SERVFAIL as usual. The client receives the bogus data. For
1253 messages that are found to be secure the AD bit is set in
1254 replies. Also logging is performed as for full validation. The
1255 default value is "no".
1256
1257 ignore-cd-flag: <yes or no>
1258 Instruct unbound to ignore the CD flag from clients and refuse
1259 to return bogus answers to them. Thus, the CD (Checking Dis‐
1260 abled) flag does not disable checking any more. This is useful
1261 if legacy (w2008) servers that set the CD flag but cannot vali‐
1262 date DNSSEC themselves are the clients, and then unbound pro‐
1263 vides them with DNSSEC protection. The default value is "no".
1264
1265 serve-expired: <yes or no>
1266 If enabled, unbound attempts to serve old responses from cache
1267 with a TTL of serve-expired-reply-ttl in the response without
1268 waiting for the actual resolution to finish. The actual resolu‐
1269 tion answer ends up in the cache later on. Default is "no".
1270
1271 serve-expired-ttl: <seconds>
1272 Limit serving of expired responses to configured seconds after
1273 expiration. 0 disables the limit. This option only applies when
1274 serve-expired is enabled. A suggested value per RFC 8767 is be‐
1275 tween 86400 (1 day) and 259200 (3 days). The default is 0.
1276
1277 serve-expired-ttl-reset: <yes or no>
1278 Set the TTL of expired records to the serve-expired-ttl value
1279 after a failed attempt to retrieve the record from upstream.
1280 This makes sure that the expired records will be served as long
1281 as there are queries for it. Default is "no".
1282
1283 serve-expired-reply-ttl: <seconds>
1284 TTL value to use when replying with expired data. If serve-ex‐
1285 pired-client-timeout is also used then it is RECOMMENDED to use
1286 30 as the value (RFC 8767). The default is 30.
1287
1288 serve-expired-client-timeout: <msec>
1289 Time in milliseconds before replying to the client with expired
1290 data. This essentially enables the serve-stale behavior as
1291 specified in RFC 8767 that first tries to resolve before immedi‐
1292 ately responding with expired data. A recommended value per RFC
1293 8767 is 1800. Setting this to 0 will disable this behavior.
1294 Default is 0.
1295
1296 serve-original-ttl: <yes or no>
1297 If enabled, unbound will always return the original TTL as re‐
1298 ceived from the upstream name server rather than the decrement‐
1299 ing TTL as stored in the cache. This feature may be useful if
1300 unbound serves as a front-end to a hidden authoritative name
1301 server. Enabling this feature does not impact cache expiry, it
1302 only changes the TTL unbound embeds in responses to queries.
1303 Note that enabling this feature implicitly disables enforcement
1304 of the configured minimum and maximum TTL, as it is assumed
1305 users who enable this feature do not want unbound to change the
1306 TTL obtained from an upstream server. Thus, the values set us‐
1307 ing cache-min-ttl and cache-max-ttl are ignored. Default is
1308 "no".
1309
1310 val-nsec3-keysize-iterations: <"list of values">
1311 List of keysize and iteration count values, separated by spaces,
1312 surrounded by quotes. Default is "1024 150 2048 150 4096 150".
1313 This determines the maximum allowed NSEC3 iteration count before
1314 a message is simply marked insecure instead of performing the
1315 many hashing iterations. The list must be in ascending order and
1316 have at least one entry. If you set it to "1024 65535" there is
1317 no restriction to NSEC3 iteration values. This table must be
1318 kept short; a very long list could cause slower operation.
1319
1320 zonemd-permissive-mode: <yes or no>
1321 If enabled the ZONEMD verification failures are only logged and
1322 do not cause the zone to be blocked and only return servfail.
1323 Useful for testing out if it works, or if the operator only
1324 wants to be notified of a problem without disrupting service.
1325 Default is no.
1326
1327 add-holddown: <seconds>
1328 Instruct the auto-trust-anchor-file probe mechanism for RFC5011
1329 autotrust updates to add new trust anchors only after they have
1330 been visible for this time. Default is 30 days as per the RFC.
1331
1332 del-holddown: <seconds>
1333 Instruct the auto-trust-anchor-file probe mechanism for RFC5011
1334 autotrust updates to remove revoked trust anchors after they
1335 have been kept in the revoked list for this long. Default is 30
1336 days as per the RFC.
1337
1338 keep-missing: <seconds>
1339 Instruct the auto-trust-anchor-file probe mechanism for RFC5011
1340 autotrust updates to remove missing trust anchors after they
1341 have been unseen for this long. This cleans up the state file
1342 if the target zone does not perform trust anchor revocation, so
1343 this makes the auto probe mechanism work with zones that perform
1344 regular (non-5011) rollovers. The default is 366 days. The
1345 value 0 does not remove missing anchors, as per the RFC.
1346
1347 permit-small-holddown: <yes or no>
1348 Debug option that allows the autotrust 5011 rollover timers to
1349 assume very small values. Default is no.
1350
1351 key-cache-size: <number>
1352 Number of bytes size of the key cache. Default is 4 megabytes.
1353 A plain number is in bytes, append 'k', 'm' or 'g' for kilo‐
1354 bytes, megabytes or gigabytes (1024*1024 bytes in a megabyte).
1355
1356 key-cache-slabs: <number>
1357 Number of slabs in the key cache. Slabs reduce lock contention
1358 by threads. Must be set to a power of 2. Setting (close) to the
1359 number of cpus is a reasonable guess.
1360
1361 neg-cache-size: <number>
1362 Number of bytes size of the aggressive negative cache. Default
1363 is 1 megabyte. A plain number is in bytes, append 'k', 'm' or
1364 'g' for kilobytes, megabytes or gigabytes (1024*1024 bytes in a
1365 megabyte).
1366
1367 unblock-lan-zones: <yes or no>
1368 Default is disabled. If enabled, then for private address
1369 space, the reverse lookups are no longer filtered. This allows
1370 unbound when running as dns service on a host where it provides
1371 service for that host, to put out all of the queries for the
1372 'lan' upstream. When enabled, only localhost, 127.0.0.1 reverse
1373 and ::1 reverse zones are configured with default local zones.
1374 Disable the option when unbound is running as a (DHCP-) DNS net‐
1375 work resolver for a group of machines, where such lookups should
1376 be filtered (RFC compliance), this also stops potential data
1377 leakage about the local network to the upstream DNS servers.
1378
1379 insecure-lan-zones: <yes or no>
1380 Default is disabled. If enabled, then reverse lookups in pri‐
1381 vate address space are not validated. This is usually required
1382 whenever unblock-lan-zones is used.
1383
1384 local-zone: <zone> <type>
1385 Configure a local zone. The type determines the answer to give
1386 if there is no match from local-data. The types are deny,
1387 refuse, static, transparent, redirect, nodefault, typetranspar‐
1388 ent, inform, inform_deny, inform_redirect, always_transparent,
1389 always_refuse, always_nxdomain, always_null, noview, and are ex‐
1390 plained below. After that the default settings are listed. Use
1391 local-data: to enter data into the local zone. Answers for local
1392 zones are authoritative DNS answers. By default the zones are
1393 class IN.
1394
1395 If you need more complicated authoritative data, with referrals,
1396 wildcards, CNAME/DNAME support, or DNSSEC authoritative service,
1397 setup a stub-zone for it as detailed in the stub zone section
1398 below.
1399
1400 deny Do not send an answer, drop the query. If there is a match
1401 from local data, the query is answered.
1402
1403 refuse
1404 Send an error message reply, with rcode REFUSED. If there is
1405 a match from local data, the query is answered.
1406
1407 static
1408 If there is a match from local data, the query is answered.
1409 Otherwise, the query is answered with nodata or nxdomain.
1410 For a negative answer a SOA is included in the answer if
1411 present as local-data for the zone apex domain.
1412
1413 transparent
1414 If there is a match from local data, the query is answered.
1415 Otherwise if the query has a different name, the query is re‐
1416 solved normally. If the query is for a name given in local‐
1417 data but no such type of data is given in localdata, then a
1418 noerror nodata answer is returned. If no local-zone is given
1419 local-data causes a transparent zone to be created by de‐
1420 fault.
1421
1422 typetransparent
1423 If there is a match from local data, the query is answered.
1424 If the query is for a different name, or for the same name
1425 but for a different type, the query is resolved normally.
1426 So, similar to transparent but types that are not listed in
1427 local data are resolved normally, so if an A record is in the
1428 local data that does not cause a nodata reply for AAAA
1429 queries.
1430
1431 redirect
1432 The query is answered from the local data for the zone name.
1433 There may be no local data beneath the zone name. This an‐
1434 swers queries for the zone, and all subdomains of the zone
1435 with the local data for the zone. It can be used to redirect
1436 a domain to return a different address record to the end
1437 user, with local-zone: "example.com." redirect and lo‐
1438 cal-data: "example.com. A 127.0.0.1" queries for www.exam‐
1439 ple.com and www.foo.example.com are redirected, so that users
1440 with web browsers cannot access sites with suffix exam‐
1441 ple.com.
1442
1443 inform
1444 The query is answered normally, same as transparent. The
1445 client IP address (@portnumber) is printed to the logfile.
1446 The log message is: timestamp, unbound-pid, info: zonename
1447 inform IP@port queryname type class. This option can be used
1448 for normal resolution, but machines looking up infected names
1449 are logged, eg. to run antivirus on them.
1450
1451 inform_deny
1452 The query is dropped, like 'deny', and logged, like 'inform'.
1453 Ie. find infected machines without answering the queries.
1454
1455 inform_redirect
1456 The query is redirected, like 'redirect', and logged, like
1457 'inform'. Ie. answer queries with fixed data and also log
1458 the machines that ask.
1459
1460 always_transparent
1461 Like transparent, but ignores local data and resolves nor‐
1462 mally.
1463
1464 always_refuse
1465 Like refuse, but ignores local data and refuses the query.
1466
1467 always_nxdomain
1468 Like static, but ignores local data and returns nxdomain for
1469 the query.
1470
1471 always_nodata
1472 Like static, but ignores local data and returns nodata for
1473 the query.
1474
1475 always_deny
1476 Like deny, but ignores local data and drops the query.
1477
1478 always_null
1479 Always returns 0.0.0.0 or ::0 for every name in the zone.
1480 Like redirect with zero data for A and AAAA. Ignores local
1481 data in the zone. Used for some block lists.
1482
1483 noview
1484 Breaks out of that view and moves towards the global local
1485 zones for answer to the query. If the view first is no,
1486 it'll resolve normally. If view first is enabled, it'll
1487 break perform that step and check the global answers. For
1488 when the view has view specific overrides but some zone has
1489 to be answered from global local zone contents.
1490
1491 nodefault
1492 Used to turn off default contents for AS112 zones. The other
1493 types also turn off default contents for the zone. The 'node‐
1494 fault' option has no other effect than turning off default
1495 contents for the given zone. Use nodefault if you use ex‐
1496 actly that zone, if you want to use a subzone, use transpar‐
1497 ent.
1498
1499 The default zones are localhost, reverse 127.0.0.1 and ::1, the onion,
1500 test, invalid and the AS112 zones. The AS112 zones are reverse DNS
1501 zones for private use and reserved IP addresses for which the servers
1502 on the internet cannot provide correct answers. They are configured by
1503 default to give nxdomain (no reverse information) answers. The defaults
1504 can be turned off by specifying your own local-zone of that name, or
1505 using the 'nodefault' type. Below is a list of the default zone con‐
1506 tents.
1507
1508 localhost
1509 The IP4 and IP6 localhost information is given. NS and SOA
1510 records are provided for completeness and to satisfy some DNS
1511 update tools. Default content:
1512 local-zone: "localhost." redirect
1513 local-data: "localhost. 10800 IN NS localhost."
1514 local-data: "localhost. 10800 IN
1515 SOA localhost. nobody.invalid. 1 3600 1200 604800 10800"
1516 local-data: "localhost. 10800 IN A 127.0.0.1"
1517 local-data: "localhost. 10800 IN AAAA ::1"
1518
1519 reverse IPv4 loopback
1520 Default content:
1521 local-zone: "127.in-addr.arpa." static
1522 local-data: "127.in-addr.arpa. 10800 IN NS localhost."
1523 local-data: "127.in-addr.arpa. 10800 IN
1524 SOA localhost. nobody.invalid. 1 3600 1200 604800 10800"
1525 local-data: "1.0.0.127.in-addr.arpa. 10800 IN
1526 PTR localhost."
1527
1528 reverse IPv6 loopback
1529 Default content:
1530 local-zone: "1.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.
1531 0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.ip6.arpa." static
1532 local-data: "1.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.
1533 0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.ip6.arpa. 10800 IN
1534 NS localhost."
1535 local-data: "1.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.
1536 0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.ip6.arpa. 10800 IN
1537 SOA localhost. nobody.invalid. 1 3600 1200 604800 10800"
1538 local-data: "1.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.
1539 0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.ip6.arpa. 10800 IN
1540 PTR localhost."
1541
1542 onion (RFC 7686)
1543 Default content:
1544 local-zone: "onion." static
1545 local-data: "onion. 10800 IN NS localhost."
1546 local-data: "onion. 10800 IN
1547 SOA localhost. nobody.invalid. 1 3600 1200 604800 10800"
1548
1549 test (RFC 6761)
1550 Default content:
1551 local-zone: "test." static
1552 local-data: "test. 10800 IN NS localhost."
1553 local-data: "test. 10800 IN
1554 SOA localhost. nobody.invalid. 1 3600 1200 604800 10800"
1555
1556 invalid (RFC 6761)
1557 Default content:
1558 local-zone: "invalid." static
1559 local-data: "invalid. 10800 IN NS localhost."
1560 local-data: "invalid. 10800 IN
1561 SOA localhost. nobody.invalid. 1 3600 1200 604800 10800"
1562
1563 reverse RFC1918 local use zones
1564 Reverse data for zones 10.in-addr.arpa, 16.172.in-addr.arpa
1565 to 31.172.in-addr.arpa, 168.192.in-addr.arpa. The lo‐
1566 cal-zone: is set static and as local-data: SOA and NS records
1567 are provided.
1568
1569 reverse RFC3330 IP4 this, link-local, testnet and broadcast
1570 Reverse data for zones 0.in-addr.arpa, 254.169.in-addr.arpa,
1571 2.0.192.in-addr.arpa (TEST NET 1), 100.51.198.in-addr.arpa
1572 (TEST NET 2), 113.0.203.in-addr.arpa (TEST NET 3),
1573 255.255.255.255.in-addr.arpa. And from 64.100.in-addr.arpa
1574 to 127.100.in-addr.arpa (Shared Address Space).
1575
1576 reverse RFC4291 IP6 unspecified
1577 Reverse data for zone
1578 0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.
1579 0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.ip6.arpa.
1580
1581 reverse RFC4193 IPv6 Locally Assigned Local Addresses
1582 Reverse data for zone D.F.ip6.arpa.
1583
1584 reverse RFC4291 IPv6 Link Local Addresses
1585 Reverse data for zones 8.E.F.ip6.arpa to B.E.F.ip6.arpa.
1586
1587 reverse IPv6 Example Prefix
1588 Reverse data for zone 8.B.D.0.1.0.0.2.ip6.arpa. This zone is
1589 used for tutorials and examples. You can remove the block on
1590 this zone with:
1591 local-zone: 8.B.D.0.1.0.0.2.ip6.arpa. nodefault
1592 You can also selectively unblock a part of the zone by making
1593 that part transparent with a local-zone statement. This also
1594 works with the other default zones.
1595
1596 local-data: "<resource record string>"
1597 Configure local data, which is served in reply to queries for it.
1598 The query has to match exactly unless you configure the local-zone
1599 as redirect. If not matched exactly, the local-zone type deter‐
1600 mines further processing. If local-data is configured that is not
1601 a subdomain of a local-zone, a transparent local-zone is config‐
1602 ured. For record types such as TXT, use single quotes, as in lo‐
1603 cal-data: 'example. TXT "text"'.
1604
1605 If you need more complicated authoritative data, with referrals,
1606 wildcards, CNAME/DNAME support, or DNSSEC authoritative service,
1607 setup a stub-zone for it as detailed in the stub zone section be‐
1608 low.
1609
1610 local-data-ptr: "IPaddr name"
1611 Configure local data shorthand for a PTR record with the reversed
1612 IPv4 or IPv6 address and the host name. For example "192.0.2.4
1613 www.example.com". TTL can be inserted like this: "2001:DB8::4
1614 7200 www.example.com"
1615
1616 local-zone-tag: <zone> <"list of tags">
1617 Assign tags to localzones. Tagged localzones will only be applied
1618 when the used access-control element has a matching tag. Tags must
1619 be defined in define-tags. Enclose list of tags in quotes ("")
1620 and put spaces between tags. When there are multiple tags it
1621 checks if the intersection of the list of tags for the query and
1622 local-zone-tag is non-empty.
1623
1624 local-zone-override: <zone> <IP netblock> <type>
1625 Override the localzone type for queries from addresses matching
1626 netblock. Use this localzone type, regardless the type configured
1627 for the local-zone (both tagged and untagged) and regardless the
1628 type configured using access-control-tag-action.
1629
1630 response-ip: <IP-netblock> <action>
1631 This requires use of the "respip" module.
1632
1633 If the IP address in an AAAA or A RR in the answer section of a
1634 response matches the specified IP netblock, the specified action
1635 will apply. <action> has generally the same semantics as that for
1636 access-control-tag-action, but there are some exceptions.
1637
1638 Actions for response-ip are different from those for local-zone in
1639 that in case of the former there is no point of such conditions as
1640 "the query matches it but there is no local data". Because of
1641 this difference, the semantics of response-ip actions are modified
1642 or simplified as follows: The static, refuse, transparent, type‐
1643 transparent, and nodefault actions are invalid for response-ip.
1644 Using any of these will cause the configuration to be rejected as
1645 faulty. The deny action is non-conditional, i.e. it always results
1646 in dropping the corresponding query. The resolution result before
1647 applying the deny action is still cached and can be used for other
1648 queries.
1649
1650 response-ip-data: <IP-netblock> <"resource record string">
1651 This requires use of the "respip" module.
1652
1653 This specifies the action data for response-ip with action being
1654 to redirect as specified by "resource record string". "Resource
1655 record string" is similar to that of access-control-tag-action,
1656 but it must be of either AAAA, A or CNAME types. If the IP-net‐
1657 block is an IPv6/IPV4 prefix, the record must be AAAA/A respec‐
1658 tively, unless it is a CNAME (which can be used for both versions
1659 of IP netblocks). If it is CNAME there must not be more than one
1660 response-ip-data for the same IP-netblock. Also, CNAME and other
1661 types of records must not coexist for the same IP-netblock, fol‐
1662 lowing the normal rules for CNAME records. The textual domain
1663 name for the CNAME does not have to be explicitly terminated with
1664 a dot ("."); the root name is assumed to be the origin for the
1665 name.
1666
1667 response-ip-tag: <IP-netblock> <"list of tags">
1668 This requires use of the "respip" module.
1669
1670 Assign tags to response IP-netblocks. If the IP address in an
1671 AAAA or A RR in the answer section of a response matches the spec‐
1672 ified IP-netblock, the specified tags are assigned to the IP ad‐
1673 dress. Then, if an access-control-tag is defined for the client
1674 and it includes one of the tags for the response IP, the corre‐
1675 sponding access-control-tag-action will apply. Tag matching rule
1676 is the same as that for access-control-tag and local-zones. Un‐
1677 like local-zone-tag, response-ip-tag can be defined for an IP-net‐
1678 block even if no response-ip is defined for that netblock. If
1679 multiple response-ip-tag options are specified for the same IP-
1680 netblock in different statements, all but the first will be ig‐
1681 nored. However, this will not be flagged as a configuration er‐
1682 ror, but the result is probably not what was intended.
1683
1684 Actions specified in an access-control-tag-action that has a
1685 matching tag with response-ip-tag can be those that are "invalid"
1686 for response-ip listed above, since access-control-tag-actions can
1687 be shared with local zones. For these actions, if they behave
1688 differently depending on whether local data exists or not in case
1689 of local zones, the behavior for response-ip-data will generally
1690 result in NOERROR/NODATA instead of NXDOMAIN, since the response-
1691 ip data are inherently type specific, and non-existence of data
1692 does not indicate anything about the existence or non-existence of
1693 the qname itself. For example, if the matching tag action is
1694 static but there is no data for the corresponding response-ip con‐
1695 figuration, then the result will be NOERROR/NODATA. The only case
1696 where NXDOMAIN is returned is when an always_nxdomain action ap‐
1697 plies.
1698
1699 ratelimit: <number or 0>
1700 Enable ratelimiting of queries sent to nameserver for performing
1701 recursion. If 0, the default, it is disabled. This option is ex‐
1702 perimental at this time. The ratelimit is in queries per second
1703 that are allowed. More queries are turned away with an error
1704 (servfail). This stops recursive floods, eg. random query names,
1705 but not spoofed reflection floods. Cached responses are not rate‐
1706 limited by this setting. The zone of the query is determined by
1707 examining the nameservers for it, the zone name is used to keep
1708 track of the rate. For example, 1000 may be a suitable value to
1709 stop the server from being overloaded with random names, and keeps
1710 unbound from sending traffic to the nameservers for those zones.
1711
1712 ratelimit-size: <memory size>
1713 Give the size of the data structure in which the current ongoing
1714 rates are kept track in. Default 4m. In bytes or use m(mega),
1715 k(kilo), g(giga). The ratelimit structure is small, so this data
1716 structure likely does not need to be large.
1717
1718 ratelimit-slabs: <number>
1719 Give power of 2 number of slabs, this is used to reduce lock con‐
1720 tention in the ratelimit tracking data structure. Close to the
1721 number of cpus is a fairly good setting.
1722
1723 ratelimit-factor: <number>
1724 Set the amount of queries to rate limit when the limit is ex‐
1725 ceeded. If set to 0, all queries are dropped for domains where
1726 the limit is exceeded. If set to another value, 1 in that number
1727 is allowed through to complete. Default is 10, allowing 1/10
1728 traffic to flow normally. This can make ordinary queries complete
1729 (if repeatedly queried for), and enter the cache, whilst also mit‐
1730 igating the traffic flow by the factor given.
1731
1732 ratelimit-for-domain: <domain> <number qps or 0>
1733 Override the global ratelimit for an exact match domain name with
1734 the listed number. You can give this for any number of names.
1735 For example, for a top-level-domain you may want to have a higher
1736 limit than other names. A value of 0 will disable ratelimiting
1737 for that domain.
1738
1739 ratelimit-below-domain: <domain> <number qps or 0>
1740 Override the global ratelimit for a domain name that ends in this
1741 name. You can give this multiple times, it then describes differ‐
1742 ent settings in different parts of the namespace. The closest
1743 matching suffix is used to determine the qps limit. The rate for
1744 the exact matching domain name is not changed, use rate‐
1745 limit-for-domain to set that, you might want to use different set‐
1746 tings for a top-level-domain and subdomains. A value of 0 will
1747 disable ratelimiting for domain names that end in this name.
1748
1749 ip-ratelimit: <number or 0>
1750 Enable global ratelimiting of queries accepted per ip address. If
1751 0, the default, it is disabled. This option is experimental at
1752 this time. The ratelimit is in queries per second that are al‐
1753 lowed. More queries are completely dropped and will not receive a
1754 reply, SERVFAIL or otherwise. IP ratelimiting happens before
1755 looking in the cache. This may be useful for mitigating amplifica‐
1756 tion attacks.
1757
1758 ip-ratelimit-size: <memory size>
1759 Give the size of the data structure in which the current ongoing
1760 rates are kept track in. Default 4m. In bytes or use m(mega),
1761 k(kilo), g(giga). The ip ratelimit structure is small, so this
1762 data structure likely does not need to be large.
1763
1764 ip-ratelimit-slabs: <number>
1765 Give power of 2 number of slabs, this is used to reduce lock con‐
1766 tention in the ip ratelimit tracking data structure. Close to the
1767 number of cpus is a fairly good setting.
1768
1769 ip-ratelimit-factor: <number>
1770 Set the amount of queries to rate limit when the limit is ex‐
1771 ceeded. If set to 0, all queries are dropped for addresses where
1772 the limit is exceeded. If set to another value, 1 in that number
1773 is allowed through to complete. Default is 10, allowing 1/10
1774 traffic to flow normally. This can make ordinary queries complete
1775 (if repeatedly queried for), and enter the cache, whilst also mit‐
1776 igating the traffic flow by the factor given.
1777
1778 fast-server-permil: <number>
1779 Specify how many times out of 1000 to pick from the set of fastest
1780 servers. 0 turns the feature off. A value of 900 would pick from
1781 the fastest servers 90 percent of the time, and would perform nor‐
1782 mal exploration of random servers for the remaining time. When
1783 prefetch is enabled (or serve-expired), such prefetches are not
1784 sped up, because there is no one waiting for it, and it presents a
1785 good moment to perform server exploration. The fast-server-num op‐
1786 tion can be used to specify the size of the fastest servers set.
1787 The default for fast-server-permil is 0.
1788
1789 fast-server-num: <number>
1790 Set the number of servers that should be used for fast server se‐
1791 lection. Only use the fastest specified number of servers with the
1792 fast-server-permil option, that turns this on or off. The default
1793 is to use the fastest 3 servers.
1794
1795 edns-client-string: <IP netblock> <string>
1796 Include an EDNS0 option containing configured ascii string in
1797 queries with destination address matching the configured IP net‐
1798 block. This configuration option can be used multiple times. The
1799 most specific match will be used.
1800
1801 edns-client-string-opcode: <opcode>
1802 EDNS0 option code for the edns-client-string option, from 0 to
1803 65535. A value from the `Reserved for Local/Experimental` range
1804 (65001-65534) should be used. Default is 65001.
1805
1806 Remote Control Options
1807 In the remote-control: clause are the declarations for the remote con‐
1808 trol facility. If this is enabled, the unbound-control(8) utility can
1809 be used to send commands to the running unbound server. The server
1810 uses these clauses to setup TLSv1 security for the connection. The un‐
1811 bound-control(8) utility also reads the remote-control section for op‐
1812 tions. To setup the correct self-signed certificates use the un‐
1813 bound-control-setup(8) utility.
1814
1815 control-enable: <yes or no>
1816 The option is used to enable remote control, default is "no". If
1817 turned off, the server does not listen for control commands.
1818
1819 control-interface: <ip address or path>
1820 Give IPv4 or IPv6 addresses or local socket path to listen on for
1821 control commands. By default localhost (127.0.0.1 and ::1) is
1822 listened to. Use 0.0.0.0 and ::0 to listen to all interfaces. If
1823 you change this and permissions have been dropped, you must
1824 restart the server for the change to take effect.
1825
1826 If you set it to an absolute path, a local socket is used. The
1827 local socket does not use the certificates and keys, so those
1828 files need not be present. To restrict access, unbound sets per‐
1829 missions on the file to the user and group that is configured, the
1830 access bits are set to allow the group members to access the con‐
1831 trol socket file. Put users that need to access the socket in the
1832 that group. To restrict access further, create a directory to put
1833 the control socket in and restrict access to that directory.
1834
1835 control-port: <port number>
1836 The port number to listen on for IPv4 or IPv6 control interfaces,
1837 default is 8953. If you change this and permissions have been
1838 dropped, you must restart the server for the change to take ef‐
1839 fect.
1840
1841 control-use-cert: <yes or no>
1842 For localhost control-interface you can disable the use of TLS by
1843 setting this option to "no", default is "yes". For local sockets,
1844 TLS is disabled and the value of this option is ignored.
1845
1846 server-key-file: <private key file>
1847 Path to the server private key, by default unbound_server.key.
1848 This file is generated by the unbound-control-setup utility. This
1849 file is used by the unbound server, but not by unbound-control.
1850
1851 server-cert-file: <certificate file.pem>
1852 Path to the server self signed certificate, by default un‐
1853 bound_server.pem. This file is generated by the unbound-con‐
1854 trol-setup utility. This file is used by the unbound server, and
1855 also by unbound-control.
1856
1857 control-key-file: <private key file>
1858 Path to the control client private key, by default unbound_con‐
1859 trol.key. This file is generated by the unbound-control-setup
1860 utility. This file is used by unbound-control.
1861
1862 control-cert-file: <certificate file.pem>
1863 Path to the control client certificate, by default unbound_con‐
1864 trol.pem. This certificate has to be signed with the server cer‐
1865 tificate. This file is generated by the unbound-control-setup
1866 utility. This file is used by unbound-control.
1867
1868 Stub Zone Options
1869 There may be multiple stub-zone: clauses. Each with a name: and zero or
1870 more hostnames or IP addresses. For the stub zone this list of name‐
1871 servers is used. Class IN is assumed. The servers should be authority
1872 servers, not recursors; unbound performs the recursive processing it‐
1873 self for stub zones.
1874
1875 The stub zone can be used to configure authoritative data to be used by
1876 the resolver that cannot be accessed using the public internet servers.
1877 This is useful for company-local data or private zones. Setup an au‐
1878 thoritative server on a different host (or different port). Enter a
1879 config entry for unbound with stub-addr: <ip address of host[@port]>.
1880 The unbound resolver can then access the data, without referring to the
1881 public internet for it.
1882
1883 This setup allows DNSSEC signed zones to be served by that authorita‐
1884 tive server, in which case a trusted key entry with the public key can
1885 be put in config, so that unbound can validate the data and set the AD
1886 bit on replies for the private zone (authoritative servers do not set
1887 the AD bit). This setup makes unbound capable of answering queries for
1888 the private zone, and can even set the AD bit ('authentic'), but the AA
1889 ('authoritative') bit is not set on these replies.
1890
1891 Consider adding server: statements for domain-insecure: and for lo‐
1892 cal-zone: name nodefault for the zone if it is a locally served zone.
1893 The insecure clause stops DNSSEC from invalidating the zone. The local
1894 zone nodefault (or transparent) clause makes the (reverse-) zone bypass
1895 unbound's filtering of RFC1918 zones.
1896
1897 name: <domain name>
1898 Name of the stub zone.
1899
1900 stub-host: <domain name>
1901 Name of stub zone nameserver. Is itself resolved before it is
1902 used.
1903
1904 stub-addr: <IP address>
1905 IP address of stub zone nameserver. Can be IP 4 or IP 6. To use
1906 a nondefault port for DNS communication append '@' with the port
1907 number. If tls is enabled, then you can append a '#' and a
1908 name, then it'll check the tls authentication certificates with
1909 that name. If you combine the '@' and '#', the '@' comes first.
1910
1911 stub-prime: <yes or no>
1912 This option is by default no. If enabled it performs NS set
1913 priming, which is similar to root hints, where it starts using
1914 the list of nameservers currently published by the zone. Thus,
1915 if the hint list is slightly outdated, the resolver picks up a
1916 correct list online.
1917
1918 stub-first: <yes or no>
1919 If enabled, a query is attempted without the stub clause if it
1920 fails. The data could not be retrieved and would have caused
1921 SERVFAIL because the servers are unreachable, instead it is
1922 tried without this clause. The default is no.
1923
1924 stub-tls-upstream: <yes or no>
1925 Enabled or disable whether the queries to this stub use TLS for
1926 transport. Default is no.
1927
1928 stub-ssl-upstream: <yes or no>
1929 Alternate syntax for stub-tls-upstream.
1930
1931 stub-no-cache: <yes or no>
1932 Default is no. If enabled, data inside the stub is not cached.
1933 This is useful when you want immediate changes to be visible.
1934
1935 Forward Zone Options
1936 There may be multiple forward-zone: clauses. Each with a name: and zero
1937 or more hostnames or IP addresses. For the forward zone this list of
1938 nameservers is used to forward the queries to. The servers listed as
1939 forward-host: and forward-addr: have to handle further recursion for
1940 the query. Thus, those servers are not authority servers, but are
1941 (just like unbound is) recursive servers too; unbound does not perform
1942 recursion itself for the forward zone, it lets the remote server do it.
1943 Class IN is assumed. CNAMEs are chased by unbound itself, asking the
1944 remote server for every name in the indirection chain, to protect the
1945 local cache from illegal indirect referenced items. A forward-zone en‐
1946 try with name "." and a forward-addr target will forward all queries to
1947 that other server (unless it can answer from the cache).
1948
1949 name: <domain name>
1950 Name of the forward zone.
1951
1952 forward-host: <domain name>
1953 Name of server to forward to. Is itself resolved before it is
1954 used.
1955
1956 forward-addr: <IP address>
1957 IP address of server to forward to. Can be IP 4 or IP 6. To use
1958 a nondefault port for DNS communication append '@' with the port
1959 number. If tls is enabled, then you can append a '#' and a
1960 name, then it'll check the tls authentication certificates with
1961 that name. If you combine the '@' and '#', the '@' comes first.
1962
1963 At high verbosity it logs the TLS certificate, with TLS enabled.
1964 If you leave out the '#' and auth name from the forward-addr,
1965 any name is accepted. The cert must also match a CA from the
1966 tls-cert-bundle.
1967
1968 forward-first: <yes or no>
1969 If a forwarded query is met with a SERVFAIL error, and this op‐
1970 tion is enabled, unbound will fall back to normal recursive res‐
1971 olution for this query as if no query forwarding had been speci‐
1972 fied. The default is "no".
1973
1974 forward-tls-upstream: <yes or no>
1975 Enabled or disable whether the queries to this forwarder use TLS
1976 for transport. Default is no. If you enable this, also config‐
1977 ure a tls-cert-bundle or use tls-win-cert to load CA certs, oth‐
1978 erwise the connections cannot be authenticated.
1979
1980 forward-ssl-upstream: <yes or no>
1981 Alternate syntax for forward-tls-upstream.
1982
1983 forward-no-cache: <yes or no>
1984 Default is no. If enabled, data inside the forward is not
1985 cached. This is useful when you want immediate changes to be
1986 visible.
1987
1988 Authority Zone Options
1989 Authority zones are configured with auth-zone:, and each one must have
1990 a name:. There can be multiple ones, by listing multiple auth-zone
1991 clauses, each with a different name, pertaining to that part of the
1992 namespace. The authority zone with the name closest to the name looked
1993 up is used. Authority zones are processed after local-zones and before
1994 cache (for-downstream: yes), and when used in this manner make unbound
1995 respond like an authority server. Authority zones are also processed
1996 after cache, just before going to the network to fetch information for
1997 recursion (for-upstream: yes), and when used in this manner provide a
1998 local copy of an authority server that speeds up lookups of that data.
1999
2000 Authority zones can be read from zonefile. And can be kept updated via
2001 AXFR and IXFR. After update the zonefile is rewritten. The update
2002 mechanism uses the SOA timer values and performs SOA UDP queries to de‐
2003 tect zone changes.
2004
2005 If the update fetch fails, the timers in the SOA record are used to
2006 time another fetch attempt. Until the SOA expiry timer is reached.
2007 Then the zone is expired. When a zone is expired, queries are SERV‐
2008 FAIL, and any new serial number is accepted from the primary (even if
2009 older), and if fallback is enabled, the fallback activates to fetch
2010 from the upstream instead of the SERVFAIL.
2011
2012 name: <zone name>
2013 Name of the authority zone.
2014
2015 primary: <IP address or host name>
2016 Where to download a copy of the zone from, with AXFR and IXFR.
2017 Multiple primaries can be specified. They are all tried if one
2018 fails. To use a nondefault port for DNS communication append
2019 '@' with the port number. You can append a '#' and a name, then
2020 AXFR over TLS can be used and the tls authentication certifi‐
2021 cates will be checked with that name. If you combine the '@'
2022 and '#', the '@' comes first. If you point it at another Un‐
2023 bound instance, it would not work because that does not support
2024 AXFR/IXFR for the zone, but if you used url: to download the
2025 zonefile as a text file from a webserver that would work. If
2026 you specify the hostname, you cannot use the domain from the
2027 zonefile, because it may not have that when retrieving that
2028 data, instead use a plain IP address to avoid a circular depen‐
2029 dency on retrieving that IP address.
2030
2031 master: <IP address or host name>
2032 Alternate syntax for primary.
2033
2034 url: <url to zonefile>
2035 Where to download a zonefile for the zone. With http or https.
2036 An example for the url is "http://www.example.com/exam‐
2037 ple.org.zone". Multiple url statements can be given, they are
2038 tried in turn. If only urls are given the SOA refresh timer is
2039 used to wait for making new downloads. If also primaries are
2040 listed, the primaries are first probed with UDP SOA queries to
2041 see if the SOA serial number has changed, reducing the number of
2042 downloads. If none of the urls work, the primaries are tried
2043 with IXFR and AXFR. For https, the tls-cert-bundle and the
2044 hostname from the url are used to authenticate the connection.
2045 If you specify a hostname in the URL, you cannot use the domain
2046 from the zonefile, because it may not have that when retrieving
2047 that data, instead use a plain IP address to avoid a circular
2048 dependency on retrieving that IP address. Avoid dependencies on
2049 name lookups by using a notation like "http://192.0.2.1/unbound-
2050 primaries/example.com.zone", with an explicit IP address.
2051
2052 allow-notify: <IP address or host name or netblockIP/prefix>
2053 With allow-notify you can specify additional sources of noti‐
2054 fies. When notified, the server attempts to first probe and
2055 then zone transfer. If the notify is from a primary, it first
2056 attempts that primary. Otherwise other primaries are attempted.
2057 If there are no primaries, but only urls, the file is downloaded
2058 when notified. The primaries from primary: statements are al‐
2059 lowed notify by default.
2060
2061 fallback-enabled: <yes or no>
2062 Default no. If enabled, unbound falls back to querying the in‐
2063 ternet as a resolver for this zone when lookups fail. For exam‐
2064 ple for DNSSEC validation failures.
2065
2066 for-downstream: <yes or no>
2067 Default yes. If enabled, unbound serves authority responses to
2068 downstream clients for this zone. This option makes unbound be‐
2069 have, for the queries with names in this zone, like one of the
2070 authority servers for that zone. Turn it off if you want un‐
2071 bound to provide recursion for the zone but have a local copy of
2072 zone data. If for-downstream is no and for-upstream is yes,
2073 then unbound will DNSSEC validate the contents of the zone be‐
2074 fore serving the zone contents to clients and store validation
2075 results in the cache.
2076
2077 for-upstream: <yes or no>
2078 Default yes. If enabled, unbound fetches data from this data
2079 collection for answering recursion queries. Instead of sending
2080 queries over the internet to the authority servers for this
2081 zone, it'll fetch the data directly from the zone data. Turn it
2082 on when you want unbound to provide recursion for downstream
2083 clients, and use the zone data as a local copy to speed up
2084 lookups.
2085
2086 zonemd-check: <yes or no>
2087 Enable this option to check ZONEMD records in the zone. Default
2088 is disabled. The ZONEMD record is a checksum over the zone
2089 data. This includes glue in the zone and data from the zone
2090 file, and excludes comments from the zone file. When there is a
2091 DNSSEC chain of trust, DNSSEC signatures are checked too.
2092
2093 zonemd-reject-absence: <yes or no>
2094 Enable this option to reject the absence of the ZONEMD record.
2095 Without it, when zonemd is not there it is not checked. It is
2096 useful to enable for a nonDNSSEC signed zone where the operator
2097 wants to require the verification of a ZONEMD, hence a missing
2098 ZONEMD is a failure. The action upon failure is controlled by
2099 the zonemd-permissive-mode option, for log only or also block
2100 the zone. The default is no.
2101
2102 Without the option absence of a ZONEMD is only a failure when
2103 the zone is DNSSEC signed, and we have a trust anchor, and the
2104 DNSSEC verification of the absence of the ZONEMD fails. With
2105 the option enabled, the absence of a ZONEMD is always a failure,
2106 also for nonDNSSEC signed zones.
2107
2108 zonefile: <filename>
2109 The filename where the zone is stored. If not given then no
2110 zonefile is used. If the file does not exist or is empty, un‐
2111 bound will attempt to fetch zone data (eg. from the primary
2112 servers).
2113
2114 View Options
2115 There may be multiple view: clauses. Each with a name: and zero or more
2116 local-zone and local-data elements. Views can also contain view-first,
2117 response-ip, response-ip-data and local-data-ptr elements. View can be
2118 mapped to requests by specifying the view name in an access-con‐
2119 trol-view element. Options from matching views will override global op‐
2120 tions. Global options will be used if no matching view is found, or
2121 when the matching view does not have the option specified.
2122
2123 name: <view name>
2124 Name of the view. Must be unique. This name is used in ac‐
2125 cess-control-view elements.
2126
2127 local-zone: <zone> <type>
2128 View specific local-zone elements. Has the same types and behav‐
2129 iour as the global local-zone elements. When there is at least
2130 one local-zone specified and view-first is no, the default lo‐
2131 cal-zones will be added to this view. Defaults can be disabled
2132 using the nodefault type. When view-first is yes or when a view
2133 does not have a local-zone, the global local-zone will be used
2134 including it's default zones.
2135
2136 local-data: "<resource record string>"
2137 View specific local-data elements. Has the same behaviour as the
2138 global local-data elements.
2139
2140 local-data-ptr: "IPaddr name"
2141 View specific local-data-ptr elements. Has the same behaviour as
2142 the global local-data-ptr elements.
2143
2144 view-first: <yes or no>
2145 If enabled, it attempts to use the global local-zone and lo‐
2146 cal-data if there is no match in the view specific options. The
2147 default is no.
2148
2149 Python Module Options
2150 The python: clause gives the settings for the python(1) script module.
2151 This module acts like the iterator and validator modules do, on queries
2152 and answers. To enable the script module it has to be compiled into
2153 the daemon, and the word "python" has to be put in the module-config:
2154 option (usually first, or between the validator and iterator). Multiple
2155 instances of the python module are supported by adding the word
2156 "python" more than once.
2157
2158 If the chroot: option is enabled, you should make sure Python's library
2159 directory structure is bind mounted in the new root environment, see
2160 mount(8). Also the python-script: path should be specified as an abso‐
2161 lute path relative to the new root, or as a relative path to the work‐
2162 ing directory.
2163
2164 python-script: <python file>
2165 The script file to load. Repeat this option for every python
2166 module instance added to the module-config: option.
2167
2168 Dynamic Library Module Options
2169 The dynlib: clause gives the settings for the dynlib module. This mod‐
2170 ule is only a very small wrapper that allows dynamic modules to be
2171 loaded on runtime instead of being compiled into the application. To
2172 enable the dynlib module it has to be compiled into the daemon, and the
2173 word "dynlib" has to be put in the module-config: option. Multiple in‐
2174 stances of dynamic libraries are supported by adding the word "dynlib"
2175 more than once.
2176
2177 The dynlib-file: path should be specified as an absolute path relative
2178 to the new path set by chroot: option, or as a relative path to the
2179 working directory.
2180
2181 dynlib-file: <dynlib file>
2182 The dynamic library file to load. Repeat this option for every
2183 dynlib module instance added to the module-config: option.
2184
2185 DNS64 Module Options
2186 The dns64 module must be configured in the module-config: "dns64 val‐
2187 idator iterator" directive and be compiled into the daemon to be en‐
2188 abled. These settings go in the server: section.
2189
2190 dns64-prefix: <IPv6 prefix>
2191 This sets the DNS64 prefix to use to synthesize AAAA records
2192 with. It must be /96 or shorter. The default prefix is
2193 64:ff9b::/96.
2194
2195 dns64-synthall: <yes or no>
2196 Debug option, default no. If enabled, synthesize all AAAA
2197 records despite the presence of actual AAAA records.
2198
2199 dns64-ignore-aaaa: <name>
2200 List domain for which the AAAA records are ignored and the A
2201 record is used by dns64 processing instead. Can be entered mul‐
2202 tiple times, list a new domain for which it applies, one per
2203 line. Applies also to names underneath the name given.
2204
2205 DNSCrypt Options
2206 The dnscrypt: clause gives the settings of the dnscrypt channel. While
2207 those options are available, they are only meaningful if unbound was
2208 compiled with --enable-dnscrypt. Currently certificate and secret/pub‐
2209 lic keys cannot be generated by unbound. You can use dnscrypt-wrapper
2210 to generate those: https://github.com/cofyc/dnscrypt-wrapper/blob/mas‐
2211 ter/README.md#usage
2212
2213 dnscrypt-enable: <yes or no>
2214 Whether or not the dnscrypt config should be enabled. You may
2215 define configuration but not activate it. The default is no.
2216
2217 dnscrypt-port: <port number>
2218 On which port should dnscrypt should be activated. Note that you
2219 should have a matching interface option defined in the server
2220 section for this port.
2221
2222 dnscrypt-provider: <provider name>
2223 The provider name to use to distribute certificates. This is of
2224 the form: 2.dnscrypt-cert.example.com.. The name MUST end with a
2225 dot.
2226
2227 dnscrypt-secret-key: <path to secret key file>
2228 Path to the time limited secret key file. This option may be
2229 specified multiple times.
2230
2231 dnscrypt-provider-cert: <path to cert file>
2232 Path to the certificate related to the dnscrypt-secret-keys.
2233 This option may be specified multiple times.
2234
2235 dnscrypt-provider-cert-rotated: <path to cert file>
2236 Path to a certificate that we should be able to serve existing
2237 connection from but do not want to advertise over
2238 dnscrypt-provider's TXT record certs distribution. A typical
2239 use case is when rotating certificates, existing clients may
2240 still use the client magic from the old cert in their queries
2241 until they fetch and update the new cert. Likewise, it would al‐
2242 low one to prime the new cert/key without distributing the new
2243 cert yet, this can be useful when using a network of servers us‐
2244 ing anycast and on which the configuration may not get updated
2245 at the exact same time. By priming the cert, the servers can
2246 handle both old and new certs traffic while distributing only
2247 one. This option may be specified multiple times.
2248
2249 dnscrypt-shared-secret-cache-size: <memory size>
2250 Give the size of the data structure in which the shared secret
2251 keys are kept in. Default 4m. In bytes or use m(mega),
2252 k(kilo), g(giga). The shared secret cache is used when a same
2253 client is making multiple queries using the same public key. It
2254 saves a substantial amount of CPU.
2255
2256 dnscrypt-shared-secret-cache-slabs: <number>
2257 Give power of 2 number of slabs, this is used to reduce lock
2258 contention in the dnscrypt shared secrets cache. Close to the
2259 number of cpus is a fairly good setting.
2260
2261 dnscrypt-nonce-cache-size: <memory size>
2262 Give the size of the data structure in which the client nonces
2263 are kept in. Default 4m. In bytes or use m(mega), k(kilo),
2264 g(giga). The nonce cache is used to prevent dnscrypt message
2265 replaying. Client nonce should be unique for any pair of client
2266 pk/server sk.
2267
2268 dnscrypt-nonce-cache-slabs: <number>
2269 Give power of 2 number of slabs, this is used to reduce lock
2270 contention in the dnscrypt nonce cache. Close to the number of
2271 cpus is a fairly good setting.
2272
2273 EDNS Client Subnet Module Options
2274 The ECS module must be configured in the module-config: "subnetcache
2275 validator iterator" directive and be compiled into the daemon to be en‐
2276 abled. These settings go in the server: section.
2277
2278 If the destination address is allowed in the configuration Unbound will
2279 add the EDNS0 option to the query containing the relevant part of the
2280 client's address. When an answer contains the ECS option the response
2281 and the option are placed in a specialized cache. If the authority in‐
2282 dicated no support, the response is stored in the regular cache.
2283
2284 Additionally, when a client includes the option in its queries, Unbound
2285 will forward the option when sending the query to addresses that are
2286 explicitly allowed in the configuration using send-client-subnet. The
2287 option will always be forwarded, regardless the allowed addresses, if
2288 client-subnet-always-forward is set to yes. In this case the lookup in
2289 the regular cache is skipped.
2290
2291 The maximum size of the ECS cache is controlled by 'msg-cache-size' in
2292 the configuration file. On top of that, for each query only 100 differ‐
2293 ent subnets are allowed to be stored for each address family. Exceeding
2294 that number, older entries will be purged from cache.
2295
2296 send-client-subnet: <IP address>
2297 Send client source address to this authority. Append /num to in‐
2298 dicate a classless delegation netblock, for example like
2299 10.2.3.4/24 or 2001::11/64. Can be given multiple times. Author‐
2300 ities not listed will not receive edns-subnet information, un‐
2301 less domain in query is specified in client-subnet-zone.
2302
2303 client-subnet-zone: <domain>
2304 Send client source address in queries for this domain and its
2305 subdomains. Can be given multiple times. Zones not listed will
2306 not receive edns-subnet information, unless hosted by authority
2307 specified in send-client-subnet.
2308
2309 client-subnet-always-forward: <yes or no>
2310 Specify whether the ECS address check (configured using
2311 send-client-subnet) is applied for all queries, even if the
2312 triggering query contains an ECS record, or only for queries for
2313 which the ECS record is generated using the querier address (and
2314 therefore did not contain ECS data in the client query). If en‐
2315 abled, the address check is skipped when the client query con‐
2316 tains an ECS record. And the lookup in the regular cache is
2317 skipped. Default is no.
2318
2319 max-client-subnet-ipv6: <number>
2320 Specifies the maximum prefix length of the client source address
2321 we are willing to expose to third parties for IPv6. Defaults to
2322 56.
2323
2324 max-client-subnet-ipv4: <number>
2325 Specifies the maximum prefix length of the client source address
2326 we are willing to expose to third parties for IPv4. Defaults to
2327 24.
2328
2329 min-client-subnet-ipv6: <number>
2330 Specifies the minimum prefix length of the IPv6 source mask we
2331 are willing to accept in queries. Shorter source masks result in
2332 REFUSED answers. Source mask of 0 is always accepted. Default is
2333 0.
2334
2335 min-client-subnet-ipv4: <number>
2336 Specifies the minimum prefix length of the IPv4 source mask we
2337 are willing to accept in queries. Shorter source masks result in
2338 REFUSED answers. Source mask of 0 is always accepted. Default is
2339 0.
2340
2341 max-ecs-tree-size-ipv4: <number>
2342 Specifies the maximum number of subnets ECS answers kept in the
2343 ECS radix tree. This number applies for each qname/qclass/qtype
2344 tuple. Defaults to 100.
2345
2346 max-ecs-tree-size-ipv6: <number>
2347 Specifies the maximum number of subnets ECS answers kept in the
2348 ECS radix tree. This number applies for each qname/qclass/qtype
2349 tuple. Defaults to 100.
2350
2351 Opportunistic IPsec Support Module Options
2352 The IPsec module must be configured in the module-config: "ipsecmod
2353 validator iterator" directive and be compiled into the daemon to be en‐
2354 abled. These settings go in the server: section.
2355
2356 When unbound receives an A/AAAA query that is not in the cache and
2357 finds a valid answer, it will withhold returning the answer and instead
2358 will generate an IPSECKEY subquery for the same domain name. If an an‐
2359 swer was found, unbound will call an external hook passing the follow‐
2360 ing arguments:
2361
2362 QNAME
2363 Domain name of the A/AAAA and IPSECKEY query. In string for‐
2364 mat.
2365
2366 IPSECKEY TTL
2367 TTL of the IPSECKEY RRset.
2368
2369 A/AAAA
2370 String of space separated IP addresses present in the A/AAAA
2371 RRset. The IP addresses are in string format.
2372
2373 IPSECKEY
2374 String of space separated IPSECKEY RDATA present in the
2375 IPSECKEY RRset. The IPSECKEY RDATA are in DNS presentation
2376 format.
2377
2378 The A/AAAA answer is then cached and returned to the client. If the
2379 external hook was called the TTL changes to ensure it doesn't surpass
2380 ipsecmod-max-ttl.
2381
2382 The same procedure is also followed when prefetch: is used, but the
2383 A/AAAA answer is given to the client before the hook is called. ipsec‐
2384 mod-max-ttl ensures that the A/AAAA answer given from cache is still
2385 relevant for opportunistic IPsec.
2386
2387 ipsecmod-enabled: <yes or no>
2388 Specifies whether the IPsec module is enabled or not. The IPsec
2389 module still needs to be defined in the module-config: direc‐
2390 tive. This option facilitates turning on/off the module without
2391 restarting/reloading unbound. Defaults to yes.
2392
2393 ipsecmod-hook: <filename>
2394 Specifies the external hook that unbound will call with sys‐
2395 tem(3). The file can be specified as an absolute/relative path.
2396 The file needs the proper permissions to be able to be executed
2397 by the same user that runs unbound. It must be present when the
2398 IPsec module is defined in the module-config: directive.
2399
2400 ipsecmod-strict: <yes or no>
2401 If enabled unbound requires the external hook to return a suc‐
2402 cess value of 0. Failing to do so unbound will reply with SERV‐
2403 FAIL. The A/AAAA answer will also not be cached. Defaults to
2404 no.
2405
2406 ipsecmod-max-ttl: <seconds>
2407 Time to live maximum for A/AAAA cached records after calling the
2408 external hook. Defaults to 3600.
2409
2410 ipsecmod-ignore-bogus: <yes or no>
2411 Specifies the behaviour of unbound when the IPSECKEY answer is
2412 bogus. If set to yes, the hook will be called and the A/AAAA
2413 answer will be returned to the client. If set to no, the hook
2414 will not be called and the answer to the A/AAAA query will be
2415 SERVFAIL. Mainly used for testing. Defaults to no.
2416
2417 ipsecmod-allow: <domain>
2418 Allow the ipsecmod functionality for the domain so that the mod‐
2419 ule logic will be executed. Can be given multiple times, for
2420 different domains. If the option is not specified, all domains
2421 are treated as being allowed (default).
2422
2423 ipsecmod-whitelist: <yes or no>
2424 Alternate syntax for ipsecmod-allow.
2425
2426 Cache DB Module Options
2427 The Cache DB module must be configured in the module-config: "validator
2428 cachedb iterator" directive and be compiled into the daemon with --en‐
2429 able-cachedb. If this module is enabled and configured, the specified
2430 backend database works as a second level cache: When Unbound cannot
2431 find an answer to a query in its built-in in-memory cache, it consults
2432 the specified backend. If it finds a valid answer in the backend, Un‐
2433 bound uses it to respond to the query without performing iterative DNS
2434 resolution. If Unbound cannot even find an answer in the backend, it
2435 resolves the query as usual, and stores the answer in the backend.
2436
2437 This module interacts with the serve-expired-* options and will reply
2438 with expired data if unbound is configured for that. Currently the use
2439 of serve-expired-client-timeout: and serve-expired-reply-ttl: is not
2440 consistent for data originating from the external cache as these will
2441 result in a reply with 0 TTL without trying to update the data first,
2442 ignoring the configured values.
2443
2444 If Unbound was built with --with-libhiredis on a system that has in‐
2445 stalled the hiredis C client library of Redis, then the "redis" backend
2446 can be used. This backend communicates with the specified Redis server
2447 over a TCP connection to store and retrieve cache data. It can be used
2448 as a persistent and/or shared cache backend. It should be noted that
2449 Unbound never removes data stored in the Redis server, even if some
2450 data have expired in terms of DNS TTL or the Redis server has cached
2451 too much data; if necessary the Redis server must be configured to
2452 limit the cache size, preferably with some kind of least-recently-used
2453 eviction policy. Additionally, the redis-expire-records option can be
2454 used in order to set the relative DNS TTL of the message as timeout to
2455 the Redis records; keep in mind that some additional memory is used per
2456 key and that the expire information is stored as absolute Unix time‐
2457 stamps in Redis (computer time must be stable). This backend uses syn‐
2458 chronous communication with the Redis server based on the assumption
2459 that the communication is stable and sufficiently fast. The thread
2460 waiting for a response from the Redis server cannot handle other DNS
2461 queries. Although the backend has the ability to reconnect to the
2462 server when the connection is closed unexpectedly and there is a con‐
2463 figurable timeout in case the server is overly slow or hangs up, these
2464 cases are assumed to be very rare. If connection close or timeout hap‐
2465 pens too often, Unbound will be effectively unusable with this backend.
2466 It's the administrator's responsibility to make the assumption hold.
2467
2468 The cachedb: clause gives custom settings of the cache DB module.
2469
2470 backend: <backend name>
2471 Specify the backend database name. The default database is the
2472 in-memory backend named "testframe", which, as the name sug‐
2473 gests, is not of any practical use. Depending on the build-time
2474 configuration, "redis" backend may also be used as described
2475 above.
2476
2477 secret-seed: <"secret string">
2478 Specify a seed to calculate a hash value from query information.
2479 This value will be used as the key of the corresponding answer
2480 for the backend database and can be customized if the hash
2481 should not be predictable operationally. If the backend data‐
2482 base is shared by multiple Unbound instances, all instances must
2483 use the same secret seed. This option defaults to "default".
2484
2485 The following cachedb otions are specific to the redis backend.
2486
2487 redis-server-host: <server address or name>
2488 The IP (either v6 or v4) address or domain name of the Redis
2489 server. In general an IP address should be specified as other‐
2490 wise Unbound will have to resolve the name of the server every
2491 time it establishes a connection to the server. This option de‐
2492 faults to "127.0.0.1".
2493
2494 redis-server-port: <port number>
2495 The TCP port number of the Redis server. This option defaults
2496 to 6379.
2497
2498 redis-timeout: <msec>
2499 The period until when Unbound waits for a response from the Re‐
2500 dis sever. If this timeout expires Unbound closes the connec‐
2501 tion, treats it as if the Redis server does not have the re‐
2502 quested data, and will try to re-establish a new connection
2503 later. This option defaults to 100 milliseconds.
2504
2505 redis-expire-records: <yes or no>
2506 If Redis record expiration is enabled. If yes, unbound sets
2507 timeout for Redis records so that Redis can evict keys that have
2508 expired automatically. If unbound is configured with serve-ex‐
2509 pired and serve-expired-ttl is 0, this option is internally re‐
2510 verted to "no". Redis SETEX support is required for this option
2511 (Redis >= 2.0.0). This option defaults to no.
2512
2513 DNSTAP Logging Options
2514 DNSTAP support, when compiled in, is enabled in the dnstap: section.
2515 This starts an extra thread (when compiled with threading) that writes
2516 the log information to the destination. If unbound is compiled without
2517 threading it does not spawn a thread, but connects per-process to the
2518 destination.
2519
2520 dnstap-enable: <yes or no>
2521 If dnstap is enabled. Default no. If yes, it connects to the
2522 dnstap server and if any of the dnstap-log-..-messages options
2523 is enabled it sends logs for those messages to the server.
2524
2525 dnstap-bidirectional: <yes or no>
2526 Use frame streams in bidirectional mode to transfer DNSTAP mes‐
2527 sages. Default is yes.
2528
2529 dnstap-socket-path: <file name>
2530 Sets the unix socket file name for connecting to the server that
2531 is listening on that socket. Default is "/etc/un‐
2532 bound/dnstap.sock".
2533
2534 dnstap-ip: <IPaddress[@port]>
2535 If "", the unix socket is used, if set with an IP address (IPv4
2536 or IPv6) that address is used to connect to the server.
2537
2538 dnstap-tls: <yes or no>
2539 Set this to use TLS to connect to the server specified in
2540 dnstap-ip. The default is yes. If set to no, TCP is used to
2541 connect to the server.
2542
2543 dnstap-tls-server-name: <name of TLS authentication>
2544 The TLS server name to authenticate the server with. Used when
2545 dnstap-tls is enabled. If "" it is ignored, default "".
2546
2547 dnstap-tls-cert-bundle: <file name of cert bundle>
2548 The pem file with certs to verify the TLS server certificate. If
2549 "" the server default cert bundle is used, or the windows cert
2550 bundle on windows. Default is "".
2551
2552 dnstap-tls-client-key-file: <file name>
2553 The client key file for TLS client authentication. If "" client
2554 authentication is not used. Default is "".
2555
2556 dnstap-tls-client-cert-file: <file name>
2557 The client cert file for TLS client authentication. Default is
2558 "".
2559
2560 dnstap-send-identity: <yes or no>
2561 If enabled, the server identity is included in the log messages.
2562 Default is no.
2563
2564 dnstap-send-version: <yes or no>
2565 If enabled, the server version if included in the log messages.
2566 Default is no.
2567
2568 dnstap-identity: <string>
2569 The identity to send with messages, if "" the hostname is used.
2570 Default is "".
2571
2572 dnstap-version: <string>
2573 The version to send with messages, if "" the package version is
2574 used. Default is "".
2575
2576 dnstap-log-resolver-query-messages: <yes or no>
2577 Enable to log resolver query messages. Default is no. These
2578 are messages from unbound to upstream servers.
2579
2580 dnstap-log-resolver-response-messages: <yes or no>
2581 Enable to log resolver response messages. Default is no. These
2582 are replies from upstream servers to unbound.
2583
2584 dnstap-log-client-query-messages: <yes or no>
2585 Enable to log client query messages. Default is no. These are
2586 client queries to unbound.
2587
2588 dnstap-log-client-response-messages: <yes or no>
2589 Enable to log client response messages. Default is no. These
2590 are responses from unbound to clients.
2591
2592 dnstap-log-forwarder-query-messages: <yes or no>
2593 Enable to log forwarder query messages. Default is no.
2594
2595 dnstap-log-forwarder-response-messages: <yes or no>
2596 Enable to log forwarder response messages. Default is no.
2597
2598 Response Policy Zone Options
2599 Response Policy Zones are configured with rpz:, and each one must have
2600 a name:. There can be multiple ones, by listing multiple rpz clauses,
2601 each with a different name. RPZ clauses are applied in order of config‐
2602 uration. The respip module needs to be added to the module-config,
2603 e.g.: module-config: "respip validator iterator".
2604
2605 Only the QNAME and Response IP Address triggers are supported. The sup‐
2606 ported RPZ actions are: NXDOMAIN, NODATA, PASSTHRU, DROP and Local
2607 Data. RPZ QNAME triggers are applied after local-zones and before auth-
2608 zones.
2609
2610 name: <zone name>
2611 Name of the authority zone.
2612
2613 primary: <IP address or host name>
2614 Where to download a copy of the zone from, with AXFR and IXFR.
2615 Multiple primaries can be specified. They are all tried if one
2616 fails. To use a nondefault port for DNS communication append
2617 '@' with the port number. You can append a '#' and a name, then
2618 AXFR over TLS can be used and the tls authentication certifi‐
2619 cates will be checked with that name. If you combine the '@'
2620 and '#', the '@' comes first. If you point it at another Un‐
2621 bound instance, it would not work because that does not support
2622 AXFR/IXFR for the zone, but if you used url: to download the
2623 zonefile as a text file from a webserver that would work. If
2624 you specify the hostname, you cannot use the domain from the
2625 zonefile, because it may not have that when retrieving that
2626 data, instead use a plain IP address to avoid a circular depen‐
2627 dency on retrieving that IP address.
2628
2629 master: <IP address or host name>
2630 Alternate syntax for primary.
2631
2632 url: <url to zonefile>
2633 Where to download a zonefile for the zone. With http or https.
2634 An example for the url is "http://www.example.com/exam‐
2635 ple.org.zone". Multiple url statements can be given, they are
2636 tried in turn. If only urls are given the SOA refresh timer is
2637 used to wait for making new downloads. If also primaries are
2638 listed, the primaries are first probed with UDP SOA queries to
2639 see if the SOA serial number has changed, reducing the number of
2640 downloads. If none of the urls work, the primaries are tried
2641 with IXFR and AXFR. For https, the tls-cert-bundle and the
2642 hostname from the url are used to authenticate the connection.
2643
2644 allow-notify: <IP address or host name or netblockIP/prefix>
2645 With allow-notify you can specify additional sources of noti‐
2646 fies. When notified, the server attempts to first probe and
2647 then zone transfer. If the notify is from a primary, it first
2648 attempts that primary. Otherwise other primaries are attempted.
2649 If there are no primaries, but only urls, the file is downloaded
2650 when notified. The primaries from primary: statements are al‐
2651 lowed notify by default.
2652
2653 zonefile: <filename>
2654 The filename where the zone is stored. If not given then no
2655 zonefile is used. If the file does not exist or is empty, un‐
2656 bound will attempt to fetch zone data (eg. from the primary
2657 servers).
2658
2659 rpz-action-override: <action>
2660 Always use this RPZ action for matching triggers from this zone.
2661 Possible action are: nxdomain, nodata, passthru, drop, disabled
2662 and cname.
2663
2664 rpz-cname-override: <domain>
2665 The CNAME target domain to use if the cname action is configured
2666 for rpz-action-override.
2667
2668 rpz-log: <yes or no>
2669 Log all applied RPZ actions for this RPZ zone. Default is no.
2670
2671 rpz-log-name: <name>
2672 Specify a string to be part of the log line, for easy referenc‐
2673 ing.
2674
2675 tags: <list of tags>
2676 Limit the policies from this RPZ clause to clients with a match‐
2677 ing tag. Tags need to be defined in define-tag and can be as‐
2678 signed to client addresses using access-control-tag. Enclose
2679 list of tags in quotes ("") and put spaces between tags. If no
2680 tags are specified the policies from this clause will be applied
2681 for all clients.
2682
2684 In the example config settings below memory usage is reduced. Some ser‐
2685 vice levels are lower, notable very large data and a high TCP load are
2686 no longer supported. Very large data and high TCP loads are exceptional
2687 for the DNS. DNSSEC validation is enabled, just add trust anchors. If
2688 you do not have to worry about programs using more than 3 Mb of memory,
2689 the below example is not for you. Use the defaults to receive full ser‐
2690 vice, which on BSD-32bit tops out at 30-40 Mb after heavy usage.
2691
2692 # example settings that reduce memory usage
2693 server:
2694 num-threads: 1
2695 outgoing-num-tcp: 1 # this limits TCP service, uses less buffers.
2696 incoming-num-tcp: 1
2697 outgoing-range: 60 # uses less memory, but less performance.
2698 msg-buffer-size: 8192 # note this limits service, 'no huge stuff'.
2699 msg-cache-size: 100k
2700 msg-cache-slabs: 1
2701 rrset-cache-size: 100k
2702 rrset-cache-slabs: 1
2703 infra-cache-numhosts: 200
2704 infra-cache-slabs: 1
2705 key-cache-size: 100k
2706 key-cache-slabs: 1
2707 neg-cache-size: 10k
2708 num-queries-per-thread: 30
2709 target-fetch-policy: "2 1 0 0 0 0"
2710 harden-large-queries: "yes"
2711 harden-short-bufsize: "yes"
2712
2714 /etc/unbound
2715 default unbound working directory.
2716
2717 /etc/unbound
2718 default chroot(2) location.
2719
2720 /etc/unbound/unbound.conf
2721 unbound configuration file.
2722
2723 /run/unbound/unbound.pid
2724 default unbound pidfile with process ID of the running daemon.
2725
2726 unbound.log
2727 unbound log file. default is to log to syslog(3).
2728
2730 unbound(8), unbound-checkconf(8).
2731
2733 Unbound was written by NLnet Labs. Please see CREDITS file in the dis‐
2734 tribution for further details.
2735
2736
2737
2738NLnet Labs Aug 12, 2021 unbound.conf(5)