You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Is there any possible workaround to set configuration in a way that will avoid entering panic?
During load balancing, Envoy will generally only consider available (healthy or degraded) hosts in an upstream cluster. However, if the percentage of available hosts in the cluster becomes too low, Envoy will disregard health status and balance either amongst all hosts or no hosts.
If you want to avoid entering panic mode, which means the upstream clusters keep in the healthy state and do not turn into unhealth. I guess maybe we just need to disable the Outlier detection ? And the upstream cluster would keep in health even they returned lots of failures.
Description
I noticed that there's an implementation difference between our MeshCircuitBreaker and Istio's (https://github.com/istio/istio/blob/7b2d8b9dacb7989d75b3b3d860988cadb9f5a848/pilot/pkg/networking/core/cluster_traffic_policy.go#L416-L423),
maybe we could add one flag to disable panic threshold and it depends on the user's choice.
The text was updated successfully, but these errors were encountered: