Daisy chaining on Internal IP

0

I am unable to get a particular Daisy Chain working in my KEMP. Here is a rough outline of the process:

public virtual service 150.1.1.1 sends all requests from a certain rule example.com back to real server 10.50.0.10:30001

private virtual service 10.50.0.10 listens on a wild card port and sends its requests back to two private lan servers:
- 10.50.1.10
- 10.50.1.11

Whenever I make a request to 'example.com', the request hangs. However if I were to go to 10.50.1.11:30001 directly, the page displays fine. I also noticed all the health checks come up okay.

From the TCPDumps, it appears that the private virtual service tries to send its information directly back out to the requestor rather than going back through the flow of things. How do I resolve this issue? Am I approaching it correctly for daisy chaining an internal virtual service IP with wildcard ports?

5 comments

Avatar
0
James Rago -- K360 Technical Product Manager

Sounds like you have transparency enabled on the First virtual service. Try disabling that.

Avatar
0
Jason Cameron

Wow, that was it. What does the transparency do exactly? What would happens (based on what i see from the client side) when i disable it?

Avatar
0
Justin Federico

Transparency passes the actual client IP to the real server. More information here: https://support.kemptechnologies.com/hc/en-us/articles/203126369-Transparency

Avatar
0
Jason Cameron

Is there any way to continue passing that on that client IP still? I would like to keep the originating client IP from the request, but obviously want it to daisy chain through my private IP for clustering.

Avatar
0
Derek Kiely

X-Forwarded-For is an option for HTTP traffic - https://support.kemptechnologies.com/hc/en-us/articles/202744899