Internal Network Failure
If a functioning K8s cluster has an internal network failure due to a disrupted server communication with the master node, BFD failure, or a K8s pod networking issue, Geo HA is triggered due to K8s dependency checks that are based on the K8s liveliness failure.
In the example shown in the following figure, the AMF or MME transparently starts using the alternate rack server. The N11/S11/S5 and N4/Sxa service addresses are migrated to site B rack B. The system continues signalling from rack B for rack A. At rack B, the session continues without any impact to existing subscriber sessions.
Note | Few in-transit calls might fail depending on the state where it is terminated before the UE re-attaches. |
