N3IWUE reconnection causes communicateWithUDM error: SYSTEM_FAILURE at AMF

Q/A
,
#1

Bug Description

N3IWUE reconnection causes communicateWithUDM error: SYSTEM_FAILURE at AMF.

To Reproduce

Steps to reproduce the behavior:

  • free5GC core is running on docker inside a VM (VirtualBox).
  • N3IWF is running in docker with other NFs.
  • N3IWUE is running in a separate VM.
  1. Run core: ‘cd ~/free5gc-compose && docker compose up -d’
  2. Modify config files with IP address and forwarding rules.
  3. Run N3IWUE: ‘~/n3iwue/run.sh’ (UE connects and ping works).
  4. Stop N3IWUE with ‘Ctrl+C’
  5. Run N3IWUE again.
  6. Errors:
  • In N3IWUE: ‘[ERRO][N3UE][IKE] Not Success’
  • In AMF: ‘[ERRO][AMF][Gmm][amf_ue_ngap_id:RU:5,AU:6(Non3GPP)][supi:SUPI:imsi-208930000001234] communicateWithUDM error: SYSTEM_FAILURE’

Screenshots

I opened a Github Issue with logs of error here

Workaround

When I restart only the UDM docker container after this error occurs, the N3IWUE connects without error. But if I reconnect the error comes again. That is the N3IWUE is able connect first time, after that I must restart UDM to connect again.

Environment

  • free5GC Version: v4.0.1
  • OS: Ubuntu 20.04 Server
  • Kernel version: 5.4.0-216-generic
  • go version: 1.21.8 linux/amd64

Note: Same Error occurred when core is run on VM without using docker.

#2

Hi @kurrshanthv,
We found a misbehavior in the response from UDM, which leads to the error you discribe. We will fix this in the later patch.

Thank you for raising this issue!
If you have any suggestion, please let us know.

#3

Thank you for your response. I am looking forward for the patch. If I find anything related to this issue I will share it here. :v: