Hi,
We have F5 VS routing traffic to a service behind OpenShift Router ( We are not using F55 CIS ). The OpenShift Route is configured as TLS Passthrough. I want to re-encrypt TLS at F5.
In case of TLS Passthrough configuration OpenShift Router determines the route based on TLS Client Hello hostname.
So I have OPS Route with host name “my-tls-passthrough-service.com” and I have F5 VS with hostname “my-f5vs.com” and a pool with single member pointing to OPS Router IP and port 443 . I have configured Client and Server SSL profiles. Also, in server SSL profile I have set “Server Name” attribute to “my-tls-passthrough-service.com” .
Everything works as expected – the request reaches the service through F5 .
The problem I have is when I configure Health Monitor. The generic HTTPS monitor doesn’t help as it checks the status of OPS Router , not the service behind it. But when I add ServerSSL profile to Health check monitor I get pool member marked down and message in local traffic log “Unable to connect “
Can you please help - without health monitor the set up is useless
In our environment we use custom HTTPS monitors where we specify the Host: header. This allows the OpenShift router to reach the application.
GET /favicon.ico HTTP/1.1\r\nHost: my-tls-passtrough-service.com\r\nConnection: Close\r\n\r\n
Hi @BramsBytes
Thanks. Yes we have done it too, but it doesn’t work. Not that I have expected it to work - when route in OpenShift is configured as TLS passthrough the routing host information is taken from TLS Client Hello Message not from HTTP Host Header
I see, and in the ServerSSL profile you did also specify the Server Name to be used for Client Hello?
Yes. Correct
And that what allowed the connection through.
So the requests are routed correctly through the VS using ServerSSL profile with host name in it. But Health checks configured with the same ServerSSL profile are failing