Implement a badRequests counter
I just tried to use the Hystrix Dashboard and it broke because of a missing rollingCountBadRequests property.
This adds that counter, and the required stuff around it.
What I'm unsure about here is how a bad request should be tracked, it's not very well documented in Hystrix itself (...or, I just couldn't find the docs). I found this issue: https://github.com/Netflix/Hystrix/issues/186 that discussed bad requests in relationship to the dash.
BREAKING CHANGE: This is a breaking change for anyone implementing custom collectors as the collector interface has a new function IncrementBadRequests().
could you share some more details about how the hystrix dashboard "broke"? what version of the dashboard? turbine?
i currently run the dashboard with no issues
It was in the templating layer and stopped all graphs except one or two, from rendering. I can check the dash version when I get to my computer. Regardless it's a hystrix metric that should be collected, I might be running of the dash master though.
Ah, sorry for not providing all info. I was running the dash directly against the application. Running it trough turbine would likely fill in the zero value and make it safe for dash consumption. Will get back to you about the exact dash version.
thanks, i'll wait to hear about your turbine investigation
Ok, I had "trying out turbine" on my summer schedule anyhow, but this is something that should be addressed regardless of whether running the stream through turbine "fixes" it or not.