Hi David, thanks for your question.

If Cx-1 fails the response will not reach any homestead at all. Diameter 
responses are hop-by-hop - they follow the same path that the request took (in 
reverse order of course). 

However if Cx-1 fails, sprout will end up retrying the request it sent to 
homestead, which will cause homestead to retry the request to the HSS. This is 
true however Cx-1 actually fails:
- If the first homestead node fails the HTTP transaction will time out on 
sprout, and it will retry to another homestead. 
- If the transport layer fails (so that Cx-1 fails but the homestead node stays 
up), homestead will reject the HTTP request, and sprout retries to another 
homestead. 

Hope this helps,
Alex.

-----Original Message-----
From: [email protected] 
[mailto:[email protected]] On Behalf Of Luong, 
David
Sent: 24 October 2014 20:05
To: [email protected]
Subject: [Clearwater] Homestead cluster and HSS

Hi PCW Team,

Let's assume there are two homestead instances in a cluster and each instance 
establishes a diameter connection with the HSS Cx interface: Cx-1 and Cx-2. If 
Cx-1 goes down after a request is sent to HSS, can the HSS send response over 
Cx-2 or does homestead expect the response to come over Cx-1?

Thank You.
David.
_______________________________________________
Clearwater mailing list
[email protected]
http://lists.projectclearwater.org/listinfo/clearwater
_______________________________________________
Clearwater mailing list
[email protected]
http://lists.projectclearwater.org/listinfo/clearwater

Reply via email to