]>
git.ipfire.org Git - thirdparty/knot-resolver.git/log
Vladimír Čunát [Wed, 18 Oct 2017 16:10:27 +0000 (18:10 +0200)]
.
Vladimír Čunát [Tue, 17 Oct 2017 11:35:58 +0000 (13:35 +0200)]
.
Positive caching seems usable now. (No xNAME, wildcards, etc.)
Problem around zone cuts:
- clear cache
- kdig fubar
- kdig DNSKEY cz
Vladimír Čunát [Mon, 16 Oct 2017 14:16:14 +0000 (16:16 +0200)]
.
Vladimír Čunát [Fri, 13 Oct 2017 15:13:50 +0000 (17:13 +0200)]
.
Vladimír Čunát [Thu, 12 Oct 2017 16:45:37 +0000 (18:45 +0200)]
.
Vladimír Čunát [Wed, 11 Oct 2017 15:52:53 +0000 (17:52 +0200)]
.
Vladimír Čunát [Mon, 9 Oct 2017 10:56:26 +0000 (12:56 +0200)]
. (last week)
Vladimír Čunát [Fri, 29 Sep 2017 16:51:59 +0000 (18:51 +0200)]
.
Vladimír Čunát [Wed, 27 Sep 2017 16:44:15 +0000 (18:44 +0200)]
.
Vladimír Čunát [Thu, 17 Aug 2017 16:23:20 +0000 (18:23 +0200)]
.
Vladimír Čunát [Wed, 27 Sep 2017 15:14:11 +0000 (17:14 +0200)]
lmdb: abstract conversion to/from knot_db_val_t
Vladimír Čunát [Thu, 21 Sep 2017 18:19:00 +0000 (20:19 +0200)]
Merge !374: release 1.4.0
Vladimír Čunát [Thu, 21 Sep 2017 15:24:35 +0000 (17:24 +0200)]
release 1.4.0 tomorrow
Vladimír Čunát [Thu, 21 Sep 2017 14:05:54 +0000 (16:05 +0200)]
Merge !373: Dockerfile: the aho-corasick module is now needed
Vladimír Čunát [Thu, 21 Sep 2017 13:57:51 +0000 (15:57 +0200)]
Dockerfile: the aho-corasick module is now needed
... even if only building
Vladimír Čunát [Thu, 21 Sep 2017 13:46:13 +0000 (15:46 +0200)]
Merge !371: FORWARD: don't validate NS in authority section
Fixes https://gitlab.labs.nic.cz/knot/knot-resolver/issues/248
Vladimír Čunát [Thu, 21 Sep 2017 11:44:58 +0000 (13:44 +0200)]
validate: wildcard expansion -> adjust to_wire
We pushed all authority to the wire, but that was unnecessary,
and in particular it clashed with not validating NS in authority when
forwarding (new change). Let's only apply this to NSEC* RRs.
Vladimír Čunát [Thu, 21 Sep 2017 13:37:10 +0000 (15:37 +0200)]
Merge !372: fixes after merge !367: root hints
Vladimír Čunát [Thu, 21 Sep 2017 09:21:55 +0000 (11:21 +0200)]
nitpick: unsigned vs. long makes a difference in printf
Sometimes I was getting huge RTT numbers in the output (>4G).
Vladimír Čunát [Thu, 21 Sep 2017 07:32:39 +0000 (09:32 +0200)]
hints: fix linkage on darwin
It's not very nice that this symbol used in hints.so is defined in
sbin/kresd, but it touches the daemon structures and thus doesn't
really belong into libkres.
Vladimír Čunát [Thu, 21 Sep 2017 08:51:37 +0000 (10:51 +0200)]
engine: nitpicks around loading root hints
"lua_" in name is not appropriate, as the function does no lua stuff
anymore, so let's unify it to "engine_".
Vladimír Čunát [Thu, 21 Sep 2017 08:09:34 +0000 (10:09 +0200)]
nitpick: unused-variable since !367 (
cac9e018c3 )
Vladimír Čunát [Wed, 20 Sep 2017 08:45:39 +0000 (10:45 +0200)]
validate: improve verbose messages
Next time it will be directly visible what record fails to validate,
which was the most time-expensive part when creating the parent commit.
Vladimír Čunát [Tue, 19 Sep 2017 13:43:03 +0000 (15:43 +0200)]
FORWARD: don't validate NS in authority section
Fixes https://gitlab.labs.nic.cz/knot/knot-resolver/issues/248
Some (exotic?) resolvers add extra NS records but doesn't provide
signatures for them even though we ask with +dnssec +cd.
That lead to validation errors. Current example server: 198.101.242.72
Let's not try to validate them when FORWARDing, as we won't most likely
need those records anyway (contrary to iteration mode).
Vladimír Čunát [Tue, 19 Sep 2017 14:15:15 +0000 (16:15 +0200)]
Merge !367: roothints: load from zonefile
Vladimír Čunát [Tue, 12 Sep 2017 13:49:58 +0000 (15:49 +0200)]
Merge !370: make: link gnutls into libkres
Vladimír Čunát [Tue, 12 Sep 2017 13:14:16 +0000 (15:14 +0200)]
make: link gnutls into libkres
It's needed after the parent merge (sometimes, e.g. on macos).
We always depended on it through libdnssec, so it's no real difference.
Vladimír Čunát [Tue, 12 Sep 2017 11:54:42 +0000 (13:54 +0200)]
Merge !369: gnutls logging improvements
Vladimír Čunát [Tue, 12 Sep 2017 11:45:30 +0000 (13:45 +0200)]
gnutls logging improvements
- move it to utils.c, so it's sensitive to later changes in verbosity
- don't mark the lines with [tls], as they may come through libdnssec
- use stdout like other verbose messages, instead of stderr (real errors)
Vladimír Čunát [Mon, 11 Sep 2017 16:38:42 +0000 (18:38 +0200)]
roothints: various changes
- expose the function as hints.root_file
- use the same filename as Debian
- remove the unneeded script
- docs and some nitpicks
Petr Špaček [Mon, 11 Sep 2017 10:55:54 +0000 (12:55 +0200)]
Merge branch 'max-resolve-time' into 'master'
Limit maximal resolution time
Closes #241
See merge request !368
Grigorii Demidov [Mon, 11 Sep 2017 08:55:19 +0000 (10:55 +0200)]
lib/resolve: optimazlization of resolution overall time checking
Vitezslav Kriz [Thu, 7 Sep 2017 14:29:18 +0000 (16:29 +0200)]
roothints: load from zonefile
Grigorii Demidov [Thu, 7 Sep 2017 13:09:24 +0000 (15:09 +0200)]
lib: additional checks when overall resolution time of single query is checked
Grigorii Demidov [Thu, 7 Sep 2017 12:07:13 +0000 (14:07 +0200)]
lib: support of an upper limit for resolution time of single query
Vladimír Čunát [Thu, 7 Sep 2017 09:11:02 +0000 (11:11 +0200)]
Merge !366: worker: disable assert we can't fix immediately
Vladimír Čunát [Thu, 7 Sep 2017 09:04:06 +0000 (11:04 +0200)]
worker: disable assertion that we can't fix immediately
Production kresd should compile with -DNDEBUG, so it wouldn't be
affected by this, but it was annoying for debugging other problems.
See https://gitlab.labs.nic.cz/knot/knot-resolver/issues/245
Vladimír Čunát [Wed, 6 Sep 2017 13:45:09 +0000 (15:45 +0200)]
Merge !362: LMDB fixes - transactions and cache.clear()
Vladimír Čunát [Wed, 6 Sep 2017 13:28:39 +0000 (15:28 +0200)]
NEWS: entry for this branch
Fixes https://gitlab.labs.nic.cz/knot/knot-resolver/issues/240
Vladimír Čunát [Mon, 4 Sep 2017 12:57:09 +0000 (14:57 +0200)]
cache: nitpicks
Vladimír Čunát [Mon, 4 Sep 2017 12:48:41 +0000 (14:48 +0200)]
cache.clear() on LMDB: be more strict about transactions
In particular, don't keep even a reset read-only transaction.
This doesn't help with that MDB_BAD_TXN, but still it seems better.
Vladimír Čunát [Mon, 4 Sep 2017 12:22:50 +0000 (14:22 +0200)]
cache.clear() on LMDB: simplify .cachelock
The locking was done incorrectly - the copied text from man open(2)
suggested creating a *unique* file and linking that one to the lockfile.
Anyway, I don't think we need to support cache on NFSv3 on old kernels ;-)
Vladimír Čunát [Mon, 4 Sep 2017 11:49:17 +0000 (13:49 +0200)]
Merge !364: doc/build: fix another instance of the old link
Vladimír Čunát [Mon, 4 Sep 2017 11:42:01 +0000 (13:42 +0200)]
doc/build: fix another instance of the old link
Vladimír Čunát [Mon, 4 Sep 2017 11:37:51 +0000 (13:37 +0200)]
Merge !363: README: Fix homepage/download link
helb [Mon, 4 Sep 2017 11:24:01 +0000 (13:24 +0200)]
Fix homepage/download link in README
Homepage link was broken – https://www.knot-resolver.cz/pages/try.html returns 404 (moved to /download during website redesign)
Vladimír Čunát [Mon, 4 Sep 2017 10:51:47 +0000 (12:51 +0200)]
cache.clear() on LMDB: first try mdb_drop()
It's certainly a cleaner way.
Vladimír Čunát [Fri, 1 Sep 2017 15:11:31 +0000 (17:11 +0200)]
cache: handle LMDB resize by another process
kresd never handled it. It's perhaps a rare occasion when different
instances (attempt to) set different cache sizes.
Vladimír Čunát [Fri, 1 Sep 2017 15:40:36 +0000 (17:40 +0200)]
docs: purge long-dead cache API from an example
Vladimír Čunát [Fri, 1 Sep 2017 13:33:30 +0000 (15:33 +0200)]
cache: rework reusing transactions (LMDB)
Previously a read transaction could be held open by each fork
indefinitely. That was done for better speed, but it had a downside
of keeping old pages alive and potentially reading only old data,
until some writes were attempted by that fork.
Now kr_cache_ provides explicit API for suitable points where to break
transactions, reusing the _sync command. On LMDB side the read-only
transaction is only reset and later renewed, supposedly giving better
performance than aborting (see LMDB docs on reset+renew).
Performance: preliminary testing with two forks, resperf on comcast
query-set shows no noticeable difference in peak QPS.
Vladimír Čunát [Fri, 1 Sep 2017 13:50:36 +0000 (15:50 +0200)]
Merge !361: main ipc_activity: misc improvements
Fixes #150.
Petr Špaček [Thu, 31 Aug 2017 11:32:53 +0000 (13:32 +0200)]
Merge branch 'ci-valgrind' into 'master'
Run unittests in CI under Valgrind memcheck
See merge request !353
Štěpán Kotek [Wed, 30 Aug 2017 09:06:11 +0000 (11:06 +0200)]
Check for memory leaks during unittest CI using valgrind
New ci task has been added that checks for possible memory leaks when running unittests. Messages from valgrind are displayed only when leak is detected.
refs #227
Štěpán Kotek [Tue, 29 Aug 2017 11:31:39 +0000 (13:31 +0200)]
Add dockerfile for valgrind ci tests.
Vitezslav Kriz [Tue, 29 Aug 2017 10:08:48 +0000 (12:08 +0200)]
main ipc_activity: close and free handle on error
Vladimír Čunát [Thu, 24 Aug 2017 11:59:01 +0000 (13:59 +0200)]
main ipc_activity: misc improvements
Stop IPC after getting an error. One point is the situation when one
of the forks ends for some reason, which lead to problems.
Another point is pipes getting out of sync.
Smaller changes:
- don't free the handle while it's still half-in-use
- don't fully panic here because of ENOMEM, just stop IPC
Fixes https://gitlab.labs.nic.cz/knot/knot-resolver/issues/150
Vladimír Čunát [Thu, 24 Aug 2017 11:50:51 +0000 (13:50 +0200)]
main ipc_activity: use goto to simplify nesting
Petr Špaček [Wed, 23 Aug 2017 15:07:52 +0000 (17:07 +0200)]
Merge branch 'ci-clone-workaround' into 'master'
Gitlab bug workaround: use git:// protocol for submodules
See merge request !360
Petr Špaček [Wed, 23 Aug 2017 14:18:02 +0000 (16:18 +0200)]
ci: update Deckard to use newest resolver test suite
Besides other things, this version of Deckard has new DNSSEC tests and
includes workaround for Gitlab CI submodule checkout.
Petr Špaček [Wed, 23 Aug 2017 13:39:23 +0000 (15:39 +0200)]
ci: use git:// protocol for external submodules
This is workaround for
https://gitlab.com/gitlab-org/gitlab-ci-multi-runner/issues/2148
Given the fact we only read it, and the fact that commit ID is stored
inside our repo, it should not bring any new security problems.
Petr Špaček [Wed, 23 Aug 2017 11:45:21 +0000 (13:45 +0200)]
Merge branch 'cache_error_reporting' into 'master'
daemon: improve error reporting related to cache configuration
See merge request !359
Petr Špaček [Tue, 22 Aug 2017 16:20:21 +0000 (18:20 +0200)]
daemon: improve error reporting related to cache configuration
Confusion related to wrong cache configuration is more frequent than it
should be. Hopefully this will enable users to help themselves.
Vladimír Čunát [Fri, 18 Aug 2017 09:26:35 +0000 (11:26 +0200)]
Merge !356: doc: fix default for net.tcp_pipeline
Vladimír Čunát [Fri, 18 Aug 2017 09:21:12 +0000 (11:21 +0200)]
doc: fix default for net.tcp_pipeline
Vladimír Čunát [Tue, 15 Aug 2017 15:21:01 +0000 (17:21 +0200)]
Merge !355: CI fixes
Vladimír Čunát [Tue, 15 Aug 2017 14:55:47 +0000 (16:55 +0200)]
gitlab-ci: employ GIT_SUBMODULE_STRATEGY: recursive
Vladimír Čunát [Tue, 15 Aug 2017 14:39:13 +0000 (16:39 +0200)]
gitlab-ci: fallout from !352
- don't have pyyaml twice
- don't install for python3 (for now), just as with other modules
Vladimír Čunát [Tue, 15 Aug 2017 14:37:47 +0000 (16:37 +0200)]
deckard: update, mainly to fix outdated RRSIG
Vladimír Čunát [Tue, 15 Aug 2017 14:21:43 +0000 (16:21 +0200)]
Merge !337: policy.suffix: update the aho-corasick code
Close https://gitlab.labs.nic.cz/knot/knot-resolver/issues/200.
Vladimír Čunát [Tue, 15 Aug 2017 14:12:12 +0000 (16:12 +0200)]
Merge branch 'master' into aho-corasick
... to resolve some simple conflicts.
Petr Špaček [Tue, 15 Aug 2017 12:09:31 +0000 (14:09 +0200)]
Merge branch 'fix-deckard-ci' into 'master'
Install yaml for deckard CI.
See merge request !352
Štěpán Kotek [Tue, 15 Aug 2017 12:04:48 +0000 (14:04 +0200)]
Install yaml for deckard CI.
Petr Špaček [Thu, 10 Aug 2017 09:58:23 +0000 (11:58 +0200)]
Merge branch 'flags-refactor' into 'master'
refactor query flags
See merge request !341
Vladimír Čunát [Wed, 9 Aug 2017 16:32:23 +0000 (18:32 +0200)]
Merge !351: Dockerfile: try fixing the build
Vladimír Čunát [Wed, 9 Aug 2017 15:28:26 +0000 (17:28 +0200)]
bootstrap-depends.sh: update packages
Vladimír Čunát [Wed, 9 Aug 2017 15:19:29 +0000 (17:19 +0200)]
Dockerfile: try fixing the build
The bad xxd didn't happen to me locally :-/
Vladimír Čunát [Wed, 9 Aug 2017 14:28:36 +0000 (16:28 +0200)]
Merge branch 'master' into flags-refactor
There were just simple conflicts in NEWS and docs.
Vladimír Čunát [Wed, 9 Aug 2017 12:19:39 +0000 (14:19 +0200)]
Merge !350: Release 1.3.3
Vladimír Čunát [Wed, 9 Aug 2017 09:27:53 +0000 (11:27 +0200)]
gitlab CI: try to fix it
Vladimír Čunát [Wed, 9 Aug 2017 08:23:47 +0000 (10:23 +0200)]
release 1.3.3
Vladimír Čunát [Wed, 9 Aug 2017 08:23:06 +0000 (10:23 +0200)]
doc: fix a RST syntax error
Vladimír Čunát [Wed, 9 Aug 2017 08:38:10 +0000 (10:38 +0200)]
Merge a critical security fix
Vladimír Čunát [Tue, 8 Aug 2017 08:26:02 +0000 (10:26 +0200)]
Merge !349: modules/http: fix compatibility with Prometheus
https://github.com/CZ-NIC/knot-resolver/pull/48
Vladimír Čunát [Tue, 8 Aug 2017 08:20:57 +0000 (10:20 +0200)]
NEWS: add parent commit
Marek Vavruša [Tue, 8 Aug 2017 05:00:32 +0000 (22:00 -0700)]
modules/http: fix compatibility with the Prometheus text format
The Prometheus text format requires histogram parameters to be quoted,
and the result to end with an empty newline characted.
Vladimír Čunát [Mon, 7 Aug 2017 08:36:02 +0000 (10:36 +0200)]
Merge !348: utils: silence a coverity warning
Vladimír Čunát [Mon, 7 Aug 2017 08:26:18 +0000 (10:26 +0200)]
utils: silence a coverity warning
There's no real change, as the function is documented to possibly
return only zero or one in this case.
Vladimír Čunát [Fri, 4 Aug 2017 12:15:30 +0000 (14:15 +0200)]
Merge !347: utils: fix seeding of the random generator
Vladimír Čunát [Fri, 4 Aug 2017 06:22:51 +0000 (08:22 +0200)]
utils: fix seeding of the random generator
If the fread didn't read all in one go, the buffer was being repeatedly
overwritten from the start instead of continuing the read :-/
I also changed the overall approach in some respects.
Vladimír Čunát [Thu, 3 Aug 2017 12:39:50 +0000 (14:39 +0200)]
dnssec: make a function static (unused outside)
Vladimír Čunát [Thu, 3 Aug 2017 11:55:04 +0000 (13:55 +0200)]
dnssec: guard against out-of-bailiwick signatures
Vladimír Čunát [Thu, 3 Aug 2017 13:23:43 +0000 (15:23 +0200)]
Merge !346: kr_rand_uint: nitpicks
Vladimír Čunát [Thu, 3 Aug 2017 12:59:48 +0000 (14:59 +0200)]
kr_rand_uint: nitpicks
- the returned value is up to max-1 and not max (tiny bias)
- improve efficiency slightly
- unsigned -> uint32_t, as that's the range it supports (mostly equal)
Vladimír Čunát [Thu, 3 Aug 2017 08:41:35 +0000 (10:41 +0200)]
Merge !343: iterate: remove counter-productive validation
Grigorii Demidov [Wed, 2 Aug 2017 16:13:49 +0000 (18:13 +0200)]
layer/iterate: remove counter-productive validation
... functionality from iterator: don't fail immediately if actual number
of labels in owner name exceeds number in label field of RRSIG rrset
Petr Špaček [Wed, 2 Aug 2017 18:45:00 +0000 (20:45 +0200)]
Merge branch 'ci-respdiff' into 'master'
CI: integrate respdiff v2 into gitlab-ci
See merge request !345
Stepan Kotek [Wed, 2 Aug 2017 10:42:01 +0000 (12:42 +0200)]
CI: migrate to respdiff v2
All resolvers are configured as recursors (no forwarding).
We should extend testing so we can test recursion and forwarding at the
same time.
List of queries is taken dynamically from
https://gitlab.labs.nic.cz/knot/knot-resolver/snippets/69
This will likely change in future.
The test will fail if mismatch rate >= 1 %.
This is error margin for network instability and few possibly broken upstreams
in the test list.
Petr Špaček [Wed, 2 Aug 2017 17:25:22 +0000 (19:25 +0200)]
Merge branch 'ci-deckard' into 'master'
CI: update Deckard to use newest resolver test suite
See merge request !344
Petr Špaček [Wed, 2 Aug 2017 17:17:42 +0000 (19:17 +0200)]
CI: update Deckard to use newest resolver test suite
Vladimír Čunát [Wed, 2 Aug 2017 15:41:51 +0000 (17:41 +0200)]
Merge !325: policy: more/better special-use domain names