Trouble Shooting

1. `ERROR: [SCTP] Failed to connect given AMF N3IWF=NGAP`

This error occured when N3IWF was started before AMF finishing initialization. This error usually appears when you run the TestNon3GPP in the first time.

Rerun the test should be fine. If it still not be solved, larger the sleeping time in line 110 of test.sh.

2. TestNon3GPP

TestNon3GPP will modify the config/amfcfg.conf. So, if you had killed the TestNon3GPP test before it finished, you might need to copy config/amfcfg.conf.bak back to config/amfcfg.conf to let other tests pass.

cp config/amfcfg.conf.bak config/amfcfg.conf

3. DB on TLS to H2C

If you meet any problems about https or mogodb, it maybe couse our new version from v3.0.1 to v3.0.2 has change http to H2C verion. Try the command below.

mongo --eval "db.NfProfile.drop()" free5gc

4. `MQCreate() Error creating message queue: Too many open files UPF=Util` (UPF)

Remove POSIX message queues

ls /dev/mqueue/
rm /dev/mqueue/*

5. Remove gtp devices (using tools in libgtp5gnl) (UPF)

cd lib/libgtp5gnl/tools
sudo ./gtp5g-link del {Dev-Name}

6. `UPF Cli Run Error: open Gtp5g: open link: create: file exists`

sudo ip link del upfgtp

7. SMF cannot communicate with UPF / UPF cannot start after a reboot or crash

The message below shows up on the logs:

[WARN][SMF][Main] Failed to setup an association with UPF[127.0.0.8], error:Request Transaction [1]: retry-out

Verify on the logs if you got this message:

[ERRO][UPF][Main] UPF Cli Run Error: open Gtp5g: open link: create: operation not supported

If yes, then try to load the GTP module on the system:

modprobe gtp5g

If this outputs an error like this:

modprobe: FATAL: Module gtp5g not found in directory /lib/modules/5.4.0-xxx-generic

Reinstall the GTP-U kernel module (refer to these instructions)

Then, once running the core run.sh script, you should obtain the message on the logs:

[INFO][SMF][Main] Received PFCP Association Setup Accepted Response from UPF[127.0.0.8]

and it should work as normal

After that, if it's required to reload the module, just run modprobe gtp5g again

Note: The symptoms described above may happen if the host machine updated it's kernel version recently

References: Free5GC Forum and Install Guide

8. Decode HTTP/2 packet in Wireshark

Run Network Function

Check has XXFsslkey.log
Edit >> Preference >> Protocols >> SSL (TLS)
Add keylog
Filter http2

9. Decode H2C (HTTP2 clear text without TLS)

The similar reason as NEA0 NAS message. Althrough H2C is clear text, wirshark still considers these packets as the normal TCP packets and does not decode them by HTTP2.

To see the details of H2C packets, do the following configuration.

Analyze → Decode As…
click Add button to add the decode rules

Decode the packets from the TCP ports listened by each NF as HTTP2 packets.

10. Clear all iptables rules

If something went wrong, it's possible to reset iptables' rules back using:

sudo iptables -P INPUT ACCEPT
sudo iptables -P FORWARD ACCEPT
sudo iptables -P OUTPUT ACCEPT
sudo iptables -t nat -F
sudo iptables -t mangle -F
sudo iptables -F
sudo iptables -X

Then remember to add back the rules required by the free5GC.

Note: You may consider using the reload_host_config.sh script for this task (see this appendix section)

11. Fix UERANSIM's EAP-AKA' behavior

If you face issues similar to the one reported on the forum or the message SEMANTICALLY_INCORRECT_MESSAGE when seting up UERANSIM's UE after configuring it to use EAP-AKA-PRIME as authentication method on Webconsole, check the solution below:

First method

Clone the nightly version from commit 85a0fbf:

Follow the cloning instructions from this guide page here

Second method

Update your UERANSIM installation with the fix:

Enter UERANSIM's folder (the folder where the source code is already cloned):
```
cd UERANSIM
```
Move the version of your source code to the commit where the fixes were merged:
```
git checkout 85a0fbf
```
Rebuild UERANSIM
```
make
```

12. N3IWUE fails to connect with `[ERRO][N3UE][IKE] Not Success` message

As per the instructions on N3IWUE install guide, the parameters of the configuration file n3ue.yaml must match those on free5GC's database.

If you get log messages like those below:

# on free5GC
[WARN][AMF][Gmm][amf_ue_ngap_id:RU:6,AU:8(Non3GPP)][supi:SUPI:imsi-208930000001234] NAS MAC verification failed(received: 0xc0c0d135, expected: 0x6f1f9365)
[ERRO][AMF][Gmm][amf_ue_ngap_id:RU:6,AU:8(Non3GPP)][supi:SUPI:imsi-208930000001234] NAS message is ciphered, but MAC verification failed
# on N3IWUE
[INFO][N3UE][IKE] Get EAP
[ERRO][N3UE][IKE] Not Success
^C4 packets captured
4 packets received by filter
0 packets dropped by kernel
[FATA][N3UE][Init] panic: runtime error: invalid memory address or nil pointer dereference

Double check if SQN matches. If not, update it on N3IWUE's side using:

nano n3iwue/config/n3ue.yaml # adjust this path if needed

Refer to the Security section on the file:

info:
    version: 1.0.1
    description: Non-3GPP UE configuration
configuration:
    N3IWFInformation:
...
    Security:
                K: b73a90cbcf3afb622dba83c58a8415df
                RAND: b120f1c1a0102a2f507dd543de68281f
                SQN: 16f3b3f71005
                AMF: 8000
                OP: b672047e003bb952dca6cb8af0e5b779
                OPC: df0c67868fa25f748b7044c6e7c245b8

Note that SQN is dinamically updated while N3IWUE is running, so this issue might get caused by the N3IWUE being killed during the authentication phase which may cause the value to not match (e.g. it was updated on just one side).

Once the parameters are updated, save and close the configuration file. Now N3IWUE should work correctly.