Commit Graph

187 Commits

Author SHA1 Message Date
Xiaotong Liu 49e3570c2c
Support server refresh concurrency (#2537)
* simultaneously refresh all servers

* Add `cert_refresh_concurrency`

---------

Co-authored-by: YX Hao <lifenjoiner@163.com>
2023-12-18 19:25:54 +08:00
keatonLiu 956a14ee21 add timeout for udp and tcp dialer 2023-12-10 23:56:13 +08:00
Frank Denis cef4b041d7 Don't call "bin" what is actually text 2023-06-24 22:11:47 +02:00
Carlo Teubner d8b1f4e7cd
Fix miscellaneous style issues (#2421)
Found by running: golangci-lint run --enable-all

I have only addressed the reported issues that seemed relevant to me.
2023-06-24 21:56:03 +02:00
lifenjoiner b3318a94b7
Try dnscrypt-proxy to resolve configured hosts when ignore_system_dns (#2204) 2023-04-05 21:17:51 +02:00
Frank Denis 361455cd58 ServiceManagerReadyNotify is not just for systemd 2022-10-20 15:33:43 +02:00
cobratbq 77059ce450
systemd: report Ready earlier as dnscrypt-proxy can itself manage retries for updates/refreshes (#2225) 2022-10-20 15:32:26 +02:00
Frank Denis 442f2e15cb Make HTTP/3 support configurable 2022-07-24 16:13:14 +02:00
Frank Denis 8c43118b03 Stop mentioning "SERVFAIL" in info messages 2022-06-19 20:38:49 +02:00
Frank Denis 5d023d2a7c Revert "New feature: sleep mode"
This reverts commit e931b234b7.
2022-04-02 09:33:49 +02:00
Frank Denis e931b234b7 New feature: sleep mode 2022-03-31 20:51:34 +02:00
Frank Denis df3fb0c9f8 Keep lines short
$ golines -w -m 120 --shorten-comments .
2022-03-23 17:48:48 +01:00
Ian Bashford 1b6caba307
allow ptr queries for cloaked domains (#1958)
* allow ptr queries for cloaked domains

* multi ips per PTR returned + cleanup

* some string tidy up

* enable config file switch

* add cloaked ptr test

* enable cloak ptrs in test scenario

* fix reverse ipv6 ptr lookup

* added ipv6 cloaked ptr test
2021-12-13 14:00:13 +01:00
Frank Denis b7704a05c5 Make return value explicit 2021-09-25 20:09:29 +02:00
Frank Denis d82021b545 Repair stale respones for DoH 2021-09-25 19:53:43 +02:00
Frank Denis 97a983c6b3 Remove leftovers from the daemonize option 2021-09-09 11:26:17 +02:00
Frank Denis 1052fa6323 serve-stale on overflow 2021-08-04 14:30:32 +02:00
Frank Denis c8a61abb79 Update comment 2021-08-04 14:27:58 +02:00
Frank Denis e64425b5e7 On overflow, only respond to cached/synthesized queries 2021-08-04 14:27:24 +02:00
Frank Denis da69583bd2 When we run out of connections, handle an extra one synchronously 2021-08-04 13:35:33 +02:00
Frank Denis d996e3424d No need to get the time if the connection is refused 2021-08-04 13:23:21 +02:00
Frank Denis 026c42424f Workaround for ODoH relays not properly forwarding the status code
Some ODoH relays return a 200 status code even when the upstream
server returns something different. This is an issue after a key
update, where a 401 code is expected.

Handle empty responses with a 200 status code as a response with
a 401 code as a workaround until these relays are fixed.
2021-07-31 12:54:23 +02:00
Frank Denis 29613096da ODoH servers should not require a static configuration 2021-06-07 13:21:58 +02:00
Frank Denis e0483bbb27 Pretend not to always use the first ODoH config 2021-06-07 12:06:36 +02:00
Frank Denis 0d81fa2796 Remove doh_client_x509_auth stuf from fetchServerInfo
It doesn't belong there, and that feature doesn't do what it's
documented to do. It sets client certificates globally instead of
doing it per server.
2021-06-07 11:23:48 +02:00
Frank Denis 27a82c54c8 ODoH: handle relay IP addresses 2021-06-07 10:46:01 +02:00
Frank Denis e57d5173e9 Support GET in ODoH targets 2021-06-06 01:22:48 +02:00
Frank Denis 44f3db31ee Just a safeguard 2021-06-05 17:57:48 +02:00
Frank Denis 0a1d3b725c Rename ODoHTarget to ODoHTargetConfig for clarity 2021-06-05 17:49:19 +02:00
Frank Denis 58e1410e66 Nits 2021-04-17 16:42:18 +02:00
Frank Denis e2e32406fb Improve ODoH log messages 2021-04-17 16:41:10 +02:00
Christopher Wood 23588733ae
Synchronously update the target configuration upon failure. (#1671)
* Synchronously update the target configuration upon failure.

* Notice a serverInfo failure when key updates fail.

* Add server name to debug logs.
2021-04-17 16:35:55 +02:00
Frank Denis 3efbacc0d4 Rename 2021-03-30 11:53:59 +02:00
Christopher Wood c748f93752 Add ODoH support. (#1653) 2021-03-30 11:53:51 +02:00
Ian Bashford fcd9225121
Threadsafe update (#1579)
* threadsafe update for relays

* locks around registeredRelays
2021-01-09 22:44:32 +01:00
Frank Denis daa1f3d3b1 Add a NOT_READY return code 2021-01-03 18:09:03 +01:00
Frank Denis 197f13ea0f Fix typo and update message 2021-01-03 16:00:02 +01:00
Frank Denis 7ec5ed127e Repair server randomization 2021-01-02 19:04:53 +01:00
Frank Denis a713e1a517 Move captive portals config to a dedicated section
Add examples
2021-01-02 15:10:04 +01:00
Frank Denis 2cdafa4bb3 Remove debug leftover 2021-01-02 10:24:32 +01:00
Frank Denis f245189f02 Handle captive portal names after coldstart 2021-01-01 21:39:17 +01:00
Ian Bashford 87fb44a588
Run from in memory cache updates (#1564)
* ConfigFile change to allowlist and blocklist

* revised names and warnings

* consistent file naming in kebab case, and generic use of blocklist and allowlist in cmoments for clarity

* update ci files

* impose maximum delay and document

* live update of servers

* update for source prefixes

* fixup test

* stop registerServers being called twice at startup

* prevent double registration at startup

* tidy function signature for loadSource

Co-authored-by: Ian Bashford <ianbashford@gmail.com>
2021-01-01 14:04:12 +01:00
Frank Denis 7f46f4820c Don't use distinct pointers for UDP and TCP relay addresses 2020-12-12 21:18:32 +01:00
Frank Denis f9c11f0897 Allow arbitrary addresses to be set in listen_addresses
Only works on OpenBSD/FreeBSD/Linux (including Android)

Fixes #1362
2020-11-25 19:23:30 +01:00
Ian Bashford 90a9a9d992
allowed ips plugin (#1510) 2020-11-15 20:59:58 +01:00
Frank Denis 4fa643ef4d Repack structures to save some memory 2020-11-14 14:46:59 +01:00
Frank Denis 272984a640 Add support for EDNS-client-subnet
Fixes #1471
2020-09-18 00:11:26 +02:00
Frank Denis 4424602e39 Start experimenting with better support for captive portals
MacOS (and probably Windows and other systems) tries to fetch a URL
before marking a network interface as available.

During this time, applications cannot use the interface at all, not
even bind their address.

When DNS queries are sent to dnscrypt-proxy, this causes the system
to wait for a response that can't come from the network, since we
hit a dead lock here.

The only option is to return hard-coded responses directly until
te interface is available.

The same captive portal configuration file can also serve a different
purpose.

Once the network is available, captive portal detection may not
work as expected if the answer is cached for too long. In fact, it
probably can't work at all since routers can't hijack DNS queries.

Once thing we can do is redirect the list of names used for captive
portal detection to the fallback resolvers. This may allow detection
to work as expected while still using a secure channel for all
other queries.
2020-08-03 18:05:42 +02:00
yofiji 7a6f1461f8
Add option to go direct for failed certificate retrieval via relay (#1397)
* Add option to go direct for failed certificate retrieval via relay

* add direct_cert_fallback to example config file

Co-authored-by: yofiji <you@example.com>
2020-07-03 12:58:36 +02:00
Frank Denis 03746b76bf Capitalize 2020-06-19 11:39:44 +02:00