-
Type: Bug
-
Resolution: Fixed
-
Priority: Major - P3
-
Affects Version/s: None
-
Component/s: None
-
Atlas Streams
-
Fully Compatible
-
Sprint 44
We have a gauge - num_ongoing_checkpoints - that is 1 when a checkpoint is in progress. But because of the way we make our metrics "visible" to the scraper (once per loop) - whenever this gauge value is scraped, at that point in the loop it is always 0 and so it seems like no checkpoints are ever ongoing in the system.
This behavior needs to be fixed