Looking for a little help around clients and poten...

# micronaut

[Mark Fisher]

Looking for a little help around clients and potential keep-alives to services.
We have a k8s environment where our product service calls our price service. We're using rolling updates in k8s to take instances out of play when there's an upgrade to either service. It's set so that there's always an instance available to serve the request.
However, during a deployment of price, we're seeing errors in the product service (Bad gateway) when trying to fetch a price, which trips the circuit and leaves us with no prices until the circuit closes again.
Is there any type of keep alive on the client trying to re-use the existing connection that might be causing this?