86028dae5a
ci(coveralls): fix glob pattern to run coveralls after merge
2022-09-20 14:24:53 -03:00
fe0ba87fd0
fix: handle interpolation of unavailable info
2022-09-20 14:15:32 -03:00
d3304d49b5
test: increase mongo integration test coverage (4.3)
2022-09-20 14:15:32 -03:00
2aaa7195f2
chore: port signing of macos binaries functionality from main-v4.3
2022-09-20 17:33:42 +02:00
f63d53e6f5
docs: update 4.3 changelog
2022-09-20 15:05:27 +02:00
b95d3bd6f8
Merge pull request #9014 from emqx/main-v4.3
...
Main v4.3 to Main v4.4 sync
2022-09-20 14:37:37 +02:00
49d1fb2b99
Merge pull request #9007 from qzhuyan/perf/william/tls-buffer-defaults-2
...
perf(TLS): set default sndbuf and recbuf to 4K
2022-09-20 08:57:10 +02:00
9262f4a862
Merge pull request #8975 from thalesmg/merge-v43-into-v44-h
...
Merge v43 into v44 h
2022-09-19 23:55:53 +02:00
bbeeafa315
Merge remote-tracking branch 'origin/main-v4.3' into merge-v43-into-v44-h
2022-09-19 20:47:45 +02:00
e7acd54645
chore: upgrade lib jose to tag emqx-1.11.3
2022-09-19 20:07:24 +02:00
94d96804b1
build(xref): delete xerf exclusion from jose_public_key
2022-09-19 19:43:07 +02:00
97fb8c1133
perf(TLS): set default sndbuf and recbuf to 4K
...
Without using default values, sockets will be opened with OS default
buffer sizes. OS default is set by linux aligning to the host memory
size that large memory has larger OS default.
Now more and more user run EMQX in container which has large OS memory
but the container max memory is limited by cgroup that could be very
small amount compared to the OS total memory size.
To eliminate uncertainty, it is better to have the default value set.
2022-09-19 18:53:43 +02:00
91dd2fc97f
Merge pull request #9005 from qzhuyan/perf/william/tls-hibernate-after-idling
...
perf(TLS): new TLS listener option to hibernate proc after idling
2022-09-19 18:47:21 +02:00
0100f210c0
chore: add new relup paths to DB
2022-09-19 15:39:18 +02:00
548de20c0b
chore: re-generate appup after version bump
2022-09-19 15:33:55 +02:00
816b4749db
chore: update app versions
2022-09-19 15:14:22 +02:00
92e1b13fa7
chore: bump version in Chart files
2022-09-19 15:12:22 +02:00
485fb5d4f0
Merge pull request #8966 from thalesmg/coveralls-test-report
...
ci: enable running test cases and cover for main-v4.x
2022-09-19 10:11:32 -03:00
044b88b8f7
chore: bump version to 4.4.10-alpha.1
2022-09-19 15:05:21 +02:00
cdadc8ad44
chore: update appup src
2022-09-19 14:53:41 +02:00
fae590b957
docs(TLS): update change log
2022-09-19 14:53:38 +02:00
cb607f760a
perf(TLS): hibernate TLS process after 5s
2022-09-19 14:53:36 +02:00
51e70df440
Merge remote-tracking branch 'origin/main-v4.3' into 0919-sync-v44
2022-09-19 09:54:00 +02:00
130e02a7af
Merge remote-tracking branch 'thales/merge-v43-into-v44-h' into 0919-sync-v44
2022-09-19 09:50:45 +02:00
35131457c0
Merge remote-tracking branch 'origin/dev/v4.4.9' into main-v4.4
2022-09-19 09:49:13 +02:00
68d28fdcdd
build(update-appup.sh): fix usage info
2022-09-18 11:32:34 +02:00
8b7726df4e
Merge pull request #8985 from zmstone/0915-sync-v4320-to-v43-trunk
...
0915 sync v4320 to v43 trunk
2022-09-18 11:10:45 +02:00
e6603548d7
chore: generate appups after bumped new version
2022-09-17 18:50:28 +02:00
49d72d24b5
Merge remote-tracking branch 'origin/main-v4.3' into 0915-sync-v4320-to-v43-trunk
2022-09-17 15:34:35 +02:00
bb2e6858b0
Merge tag 'v4.3.20' into main-v4.3
2022-09-17 14:52:03 +02:00
b7fe87d761
ci: enable running test cases and cover for main-v4.x
...
An attempt to make Coveralls bot to identify and report coverage
changes in `main-v4.{3,4}`-based PRs.
Currently, `master`-based PRs get their reports correctly. When one
checks such a report, e.g.: https://coveralls.io/builds/52535695 , it
identifies if coverage increased or decreased relative to the base
branch.
`main-v4.x`-based PRs, however, always seem to report as "first
builds", so it doesn't report if coverage increased or
decreased. E.g.: https://coveralls.io/builds/52541065 .
The reason _could be_ that, currently, `main-v4.x` branches do not
build themselves. That is, PRs run the test + cover analysis, as
usual, but once they're merged, tests (and thus coverage analysis) do
not run. And that _could be_ why the tool can't see if coverage
increased or decreased with such PR.
2022-09-16 14:18:03 -03:00
9e1d70e0cb
Merge pull request #8987 from zmstone/0916-merge-v4320-to-v449
...
0916 merge v4320 to v449
2022-09-16 17:14:46 +02:00
68dadf4ea8
chore: fix xref errors
2022-09-16 09:31:14 -03:00
2251ee715e
Merge tag 'v4.3.20' into 0915-sync-v4320-to-v43-trunk
2022-09-16 13:40:55 +02:00
f6558f897f
Merge tag 'v4.3.20' into 0916-merge-v4320-to-v449
2022-09-16 13:36:20 +02:00
8893801910
chore: bump release version to v4.3.20
2022-09-16 13:33:17 +02:00
40a0703bf5
Merge tag 'v4.3.20' into dev/v4.4.9
2022-09-16 13:11:46 +02:00
ccc47719cd
chore: bump Chart versions to 4.3.21
2022-09-16 12:58:02 +02:00
a01d01ac19
chore: bump release version to 4.3.21
2022-09-16 12:57:20 +02:00
bfb53b7f24
chore: bump app versions after merged tag v4.3.20
2022-09-16 12:56:51 +02:00
3c852098d7
Merge tag 'v4.3.20' into main-v4.3
2022-09-16 12:54:00 +02:00
bf8e3c5061
Merge pull request #8984 from HJianBo/fix-jwt-error
...
chore: avoid generating crash logs for hook execution
2022-09-16 12:38:53 +02:00
135a9a30d9
Merge pull request #8976 from thalesmg/fix-flaky-mongo-test-43
...
test: avoid inter-suite flakiness
2022-09-16 12:35:32 +02:00
89ed4a350f
chore: avoid generating crash logs for hook execution
2022-09-16 17:53:49 +08:00
f717cc9d81
Merge pull request #8983 from emqx/build-sign-macos-binaries
...
Build sign macos binaries
2022-09-16 11:01:26 +02:00
399b3f9cf9
Merge pull request #8980 from zhongwencool/fix-dashboard-bind
...
fix: add ip to dashboard listener_on
2022-09-16 10:52:53 +02:00
d02f483035
build: fix make clean
2022-09-16 10:22:08 +02:00
ed34108644
build: sign macos binaries
2022-09-16 10:21:56 +02:00
e4c2715842
fix: add ip to dashboard listener_on
2022-09-16 15:36:17 +08:00
6204481d71
test: avoid inter-suite flakiness
...
This test runs most of the time fine in CI. But, if run alone
locally, will fail consistently because the default `acl.conf` has a
catch-all `{allow, all}` clause. Probably another suite that runs
before this in CI unloads that and everything seems fine.
2022-09-15 18:08:20 -03:00
Thales Macedo Garitezi