This PR was meant to reproduce a memory leak / latency leak
with 512kb payloads, however it does not reproduce the issue.
I've refactored the time-series test into a re-usable
TimeSeriesCluster() helper. I've added three cases for 4kb,
48kb and 512kb requests.
These are tests that take around 3 minutes and are not meant
to be part of the test suite but instead run as re-usable
recipes to verify how Hyperbahn operates under load.
This PR was meant to reproduce a memory leak / latency leak with 512kb payloads, however it does not reproduce the issue.
I've refactored the time-series test into a re-usable TimeSeriesCluster() helper. I've added three cases for 4kb, 48kb and 512kb requests.
These are tests that take around 3 minutes and are not meant to be part of the test suite but instead run as re-usable recipes to verify how Hyperbahn operates under load.
r: @rf @kriskowal