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
I could not find my question answer in the repository. Can we use a startup parameter in an ingress rule? for example, I would like to use this disable-http-keepalives for only below ingress rule. I can disable it from Kubernetes skipper deployment but If I can I would like to disable keep alive for only this ingress rule not for all the k8s env.
I put it under annotations but it did not work there.
currently, this is not possible, because we use a shared connection pool and the individual routes currently don't have control on the connection states. But it's a good idea for a new feature.
How I would imagine, we could create a filter that instruments the proxy mechanism to use the connections in a specific way, like always create a new one and close after each request that are handled by a certain route. I don't know currently if this is feasible, but let's keep the issue open, maybe you or somebody else finds a way. PR welcome!
Note that we generate route objects from the ingresses, and the proxy component only deals with route objects.
@aryszka what do you think about connection pool per backend?
We could have a map[backend]*http.Client for example. Unclear to me if people would like to change it even more fine granular like per route.
Hi,
I could not find my question answer in the repository. Can we use a startup parameter in an ingress rule? for example, I would like to use this disable-http-keepalives for only below ingress rule. I can disable it from Kubernetes skipper deployment but If I can I would like to disable keep alive for only this ingress rule not for all the k8s env.
I put it under annotations but it did not work there.
The text was updated successfully, but these errors were encountered: