Spencer Sevilla
Spencer Sevilla
These edits are pretty tiny, and in-line with 3GPP: When sending a PFCP Session-Establishment-Request, header SEID should always be zero. The GTP line-edit avoids a segfault on the off-chance that...
This is a straightforward PR that makes PFCP more idempotent with respect to session deletions. Right now, if the SMF (or SGWC) sends a Session-Deletion-Request for a session that the...
OGS_GTP2_CAUSE_UE_ALREADY_RE_ATTACHED is an okay situation and does not indicate an error. The only time it is seen is when the system gracefully handles a race condition when sgwc sends downlink-data-notification...
ogs_pfcp_sendto can encounter a socket error for many short-lived and recoverable reasons, e.g. a network config changes, a link goes down, a VPN is offline, or a firewall rule comes...
sgwc_sess_remove and smf_sess_remove both have some potential segfault conditions due to double-frees. Each function will definitely segfault if called twice on the same sess structure. This is unlikely, but can...
Currently the upf/sgwc logic is a bit unpredictable when no_pfcp_rr_select = true. It goes through the list searching for an acceptable match, but if none is found it simply returns...
Right now the timeout/retransmit values for all message types (GTP, PFCP, SBI) are set to message.duration (default value 10sec). Having all these timers set to the same value can create...
Currently, if a PFCP timeout occurs (at either sgwc or smf) an error is logged but no further logic/action happens. This PR extends those timeout functions to send a GTP...
The sgwc_sess_remove() and smf_sess_remove() functions are good entry-point functions used to cleanup and remove a session. However, if they are called on an already freed session, they will segfault because...
One of the communities outside Oaxaca apparently uses a monthly membership model, and I think it'd be a good feature to add to haulage. Generally speaking I don't want to...