No queued retry on bad data #91
Conversation
Allows the queued-retry to identify permanent errors from the consumer attached to it. This way it can safely drop batches of data data without burdening the system with retries.
pjanotti
force-pushed
the
no-queued-retry-on-bad-data
branch
from
2d09dd1 to
33ad4ac
Compare
Codecov Report
@@ Coverage Diff @@
## master #91 +/- ##
==========================================
- Coverage 69.76% 69.75% -0.01%
==========================================
Files 100 101 +1
Lines 6357 6395 +38
==========================================
+ Hits 4435 4461 +26
- Misses 1692 1699 +7
- Partials 230 235 +5
Continue to review full report at Codecov.
|
You may want to increase test coverage, looks pretty low for this diff. |
|
That is unexpected... not sure if it is related to update the branch or not... at https://codecov.io/gh/open-telemetry/opentelemetry-service/pull/91?src=pr&el=continue it mentions that is unable to compare... let me see what I can figure out... |
May be a codecov bug. |
|
Build hit #59 ... updating branch... |
|
Merged due to the slowness on Travis queue. I will check the coverage numbers from other PRs. |
Previously:
TestNewInterceptor_endToEnd
served the purpose of demoing how to make an "agent"
with the OpenCensus interceptor. It just spams stdout
with output and doesn't perform any assertions
nor serve any other purpose other than demoing.
As we add other interceptors, we'll put them in
here. Instead make it an example test that compiles
and will serve as the blueprint towards making
cmd/ocagent modifications.
* b3 propagator * fix as per new api. * add benchmark test.
Allows the queued-retry to identify permanent errors from the consumer attached to it. This way it can safely drop batches of data data without burdening the system with retries.