I have an immediate requirement of enabling Ingress Traffic for a potential client. Your help is appreciated for below questions as I am about to start a POC to prove it my requirements.
1. nonstandard port communication between services across projects within in cluster and across multiple clusters.
2. Clusters running across data centers
3. routable IP’s required to reach services
4. am not talking about router or 80/443 or web traffic. As we all knew router only support standard ports http/https/TLS/SNI
am thinking to leverage OSE 3.3 ingress support to prove it in POC environment.
1. borrow routable IPs from my network team and assign to ExternalIPNetworkCIDRs in /etc/origin/master-config.yaml
2. restart the master
3. create a service and assign like below example from documentation
- name: db
4. test assigned IP by using TCP based applications.
1. is above steps enough to test?
2. What are possible problems I might enouter?
3. Is ingress feature fit into above use case?
4. How to assign pooled IP’s automatically when client requested ingress service or they need routable IP address? As per above example, it seems to be manual.
5. For my tesing, can I use few IP address instead of ExternalIPNetworkCIDRs in ExternalIPNetworkCIDRs or is it ok to have ExternalIPNetworkCIDRs in ExternalIPNetworkCIDRs but play with few ip address with in this range while creating ingress service?
6. Ingress feature works for only IPV4 or IPV6 also? I don’t have immediate requirement of IPV6 but just asking to know.
7. How ARP or MAC address works in Ingress? Will they cause any issues during production traffic or we can simply ignore? Am interested more about routing, multi cluster and multi data center scenarios …