tokers commented on issue #2739: URL: https://github.com/apache/apisix/issues/2739#issuecomment-731577350
> @gewei2000 do you have fixed this issue? find out the reason? > > what is it? The active health check target is added lazily when the upstream was first referenced, at that point, no health check information, so unhealthy endpoints might be used, especially when their weight are high or because the directional selection like consistent hash. Or after the health checker resume these offline endpoints from unhealthy to healthy, they are offline again, before the health checker spawns probe requests and achieve the threshold, these offline (or unstable) endpoints could be selected. ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected]
