networkedsystemsiitb / nfv_lte_epc Goto Github PK

Hi,
I can deploy this vEPC service, and execute the Data Traffic RAN test. The ran starts the iperf threads and the streams are received at the sink side. However at the end of the test, the PGW crashes:

pgw_handledownlinkudata: downlink udata forwarded to sgw: 56: 119000000000
pgw_handledetach:  :zero imsi 0 25: 0
0: Zero imsi: pgw_handledetach: Success
Segmentation fault (core dumped)

The SGW however seems to have received a valid imsi value from the PGW upon detach...

sgw_handledetach: detach response received from pgw: 119000000000
sgw_handledetach: pgw detach failure: 119000000000

It seems somehow the PGW uses the wrong method to get the imsi in handle_detach?

Hi,
i installed version 2 of the EPC in a private Openstack deployment and i am facing problems with the control plane traffic simulation.
I checked all the instances components and each of them is correctly up and running. When i start the control plane traffic simulation, i checked with tcpdump that traffic is correctly flowing from mme's loadbalancer to the two instances of mme i have.
However, the control plane traffic simulation is hanging due to a socket time-out error:
i checked on sctp_client.ccp source code and it tries to open a socket with mme's loadbalancer on port 5000, which is the port mme instances are listening on. Actually mme's loadbalabncer is not listening on port 5000 so the socket can not be opened. If i try to force it opening a socket with first mme's instance, the socket is correctly opened.

Do you have any clues about how to solve this problem?

Thank you in advance.