SCTP Connectivity Issue while connecting two clusters using submariner

[MuskanRajpal]

I am currently experiencing an issue with SCTP connectivity between two Kubernetes clusters while utilizing Submariner for multi-cluster communication.

Environment Details:
Clusters: Two separate Kubernetes clusters, each hosting a distinct service.
Communication: The service in the first cluster communicates with the service in the second cluster using SCTP.

Observations:
TCP Connectivity: Successful; connections are functioning as expected over TCP.
SCTP Connectivity: Encountering a connection timeout error when attempting to establish SCTP communication between the clusters.
Service Access: Services are being accessed using their respective service names.

Additional Information:

• I have tested both Calico and Flannel as the CNI (Container Network Interface) plugins.
• I am not utilizing Globalnet in this setup.
• The same use case has been tested in a single cluster environment, where SCTP communication worked perfectly, indicating that the issue arises specifically in the multi-cluster configuration.

"subctl diagnose all" output

For testing the SCTP connectivity, we are using the sctp_test command:
server:
sctp_test -H local_addr -P local_port -l
client:
sctp_test -H local_addr -P local_port -h remote_addr -p remote-port -s

I would appreciate any insights or guidance on resolving this SCTP connectivity issue.

[dfarrell07]

(note that there are ongoing discussions about this in Slack)

https://kubernetes.slack.com/archives/C010RJV694M/p1729500963229669