kafkacat has been renamed to kcat, as it looks for trademark issues.
Kafka module tests depend on that utility and are skipped if command
is not available. This was now always the case for newer development
containers.
Name was now adjusted in all tests.
- uses http library to provide http input.
user would need to configure an 'endpoint' as input, along
with a ruleset, defining how the input should be routed in
rsyslog.
bugfix, free dynamic buf if created, don't echo to client data.
do data framing using newline character.
more input options support:
- input name
- flowcontrol
add support for gzip content support, and other options
- gzip content
- parse linefeeds by default, option to ignore linefeeds
add support for gzip content support, and other options
- gzip content
- parse linefeeds by default, option to ignore linefeeds
WIP - misc edits enable port, documentroot module parameter
fix overloaded stack issue
update tests to use available port instead of hard-coded one.
Add 'octet counted framing' support
- option is "SupportOctetCountedFraming", currently 'off' by
default.
update imhttp-getrequest-file.sh test to use $srcdir
imhttp - support multi-threaded connection contexts.
- tests for large data posts
- Add header data as metadata option
move mg_start into activatecnf instead of in runinput
- plugin will get (docker) container logs from a host as well as filling out some
basic container metadata as id, name, image, labels.
- requirements: curl-7.40.0+, for unix_domain_socket option (Docker API listens on a
local socket). Additionally, plugin will handle the issue of log lines larger than
16KB begin split by Docker.
- include imdocker-unittests in testbench
- enable imdocker tests in osx travis tests
- use curl master branch for imdocker-tests
- fix cflags for pthread
- enforce minimum curl version during configuration - to 7.40.0 for imdocker
- container polls after the initial one are automatically filtered utilizing the 'since' option as per docker api
- imdocker tests refactored.
- add escapeLF config option
Test bench test added:
- imdocker-basic: checks for completeness.
- imdocker-basic-vg: same as basic with valgrind enabled.
- imdocker-long-logline: checks imdocker can handle 16K+ log lines.
- imdocker-long-logline-vg: same as long-logline test with valgrind enabled.
- add option 'retrieveNewLogsFromStart' which will cause imdocker to ignore the 'tail'
option when retrieving container logs for newly activated containers. Containers that
were already active when imdocker first starts still uses the tail option.
- imdocker-new-logs-from-start: checks that new containers will ignore the tail option
- imdocker-new-logs-from-start-vg: valgrind enabled
- multi-line support via docker label: imdocker.startregex
added multi-line tests for testbench
Travis test bench related history and notes:
- Travis tests, enable imdocker
- imdocker travis tests not enabled in precise or trusty ubuntu versions of travis-test.
Assuming travis tests will run in ubuntu 16
- tests: harden container name against strange file system pathes
RSYSLOG_DYNNAME contains the file system path, which is NOT guarnateed
to be suitable for things other than file names.
- tests: bugfix and guard against false negative
- content_check_with count was done before rsyslog was terminated, so
it was uncertain how many messages were written at that spot
(especially on slow machines)
- so far, only the number of lines was checked, not their content. So
any problems with the actual messages would not have been detected.
- improved support for listcontainer options, fix racy imdocker tests
The configure/Makefile checks were not correct, leading to the
build of journal components when not necessary, even if not
supported by the platform. Thus lead to invald build and test
failures. This commit fixes that.
This often causes trouble when the packages are rebuild by the 0mq project
(which happens frequently). We already do intensive testing of the 0mq
components in the buildbot infrastructure, where we use dedicated containers.
This is reliable, as the containers already contain everything needed and so
do not need to reach out to the 0mq package archives. In the light of this,
let's save us the trouble of Travis failures. The only downside is that
users cannot pre-test with their local Travis when modifying 0mq modules,
which is quite acceptable.
Fixed some issues with sndrcv kafka tests.
Generating kafka topics dynamically now it kafka tests.
Limited messagecount in some tests to 50000 for now.
We already know they are currently unstable, so there is not point
in messing up CI result each time. The test will still be carried out
in nightly checks, so the problems won't be forgotten.
the very new version introduced new warning compiler flags, which require
some work to fix build issues. We do not want to hold the rest of the
useful work in this commit set just for that. So we temporarily disable
the new package and will work in parallel to make things build cleanly
again.
We execute different code pathes under cron jobs. It doesn't make sense
to redo what we already did before a merge. Right now, cron builds a
Coverity scan submission and uploads it. This should be set to run
once a day.
closes https://github.com/rsyslog/rsyslog/issues/2037
this was needed while libfastjson 0.99.7 was not released
We just uncomment it so that the override can be easily re-enabled
the next time a similar need arises (what routinely happens)
closes https://github.com/rsyslog/rsyslog/issues/1848
Also fixed tryresume and doAction handling when action is
in suspend state.
Calling for Callbacks more often.
Adjusted kafka fail test settings.
diag.sh script detects of kafka server comes up proberly
Added liblz4 to dependencies for omkafka/imkafka needed
for static linking.
This addresses the omkafka queue problem from issue:
https://github.com/rsyslog/rsyslog/issues/1052
Yet it does not fully solve the message problem yet,
but I am working on this.
Package libqpid-proton3-dev is no longer available for precise.
As such, we cannot install it and so we disable both the installation
and related tests when running on precise.
sometimes the travis build complains about inability to verify
correctness, which is probably related to ongoing build work. This
should solve the issue with PPAs (mostly our own).
