Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
mainly LGTM, would be nice to clean the time constants a bit before merging
static constexpr std::chrono::milliseconds INFREQUENT_PUBLISH_PERIOD = 2 * | ||
test_constants::TEST_DURATION; |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
adding a comment to explain why this value is OK would be helpful, not sure if I understand very well what all those constants mean, and why this is safe and this test won't be flaky.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
also, it's unclear what's the value here there. Once every 10ms, 10 seconds? INFREQUENT is not really helpful in this context.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@thomas-moulard and @dabonnie, can you take a look again?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks @mm318 - before merging could you run this with --gtest_repeat=100000
to make sure it's not flaky?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I was able to repeat the test 14000+ times without fail before running into some unrelated RTPS error (like this aliasrobotics/aztarna#41)...
I am confident that the test is not flaky, as we should absolutely not see any publishing if the publishing period is set to twice that of the entire allotted test duration, even at the millisecond scale.
If we do see some publishing occur given this much margin (of the entire test duration), I would be concerned about the ROS2 executor.
Codecov Report
@@ Coverage Diff @@
## master #31 +/- ##
==========================================
- Coverage 42.04% 41.66% -0.38%
==========================================
Files 26 26
Lines 880 876 -4
Branches 506 504 -2
==========================================
- Hits 370 365 -5
- Misses 50 52 +2
+ Partials 460 459 -1
Continue to review full report at Codecov.
|
a2909d4
to
1e25c44
Compare
As alluded to at #16 (comment), this change removes the
clear_measurements_on_publish
argument fromPeriodicMeasurementNode
, treating it as if it was alwaystrue
.