Can anybody help on the setup of Temporal frontend exposed with emissary ingress and GCP http(s) LB?

I am trying to setup Temporal having frontend service exposed via GCP https load balancer and emissary ingress. Could anyone help on emissary mappings and other configs for temporal frontend service?


Is emissary / ambassador ingress a must have for your deployment? Asked our server guys about it and they mentioned they ran into a number of issues with emissary in the past. They recommended to use nginx or istio for a prod setup instead if possible.

I’m not a pro at emissary but some of the things that typically come up with load balancing temporal frontend services are
LBs should enable / support HTTP2, should set a connection timeout to 70+ seconds (to not time out your worker long-poll requests) and in in cases you configure TLS, have TLS passthrough.

Hope this helps

Thanks Tihomir. I am going to try with nginx ingress. Could you please provide any documentation/links for nginx ingress config(for grpc) and if there are any configurations at http(s) load balancer?

Just for references, here is nginx config I use locally (services running on docker currently, working on k8s sample too), just note yes it works but by no means optimized for production. Hope it helps.

moved to nginx ingress controller in kubernetes. Thank you.