'node_timeout' and some other functions can remove a smf_sess_t
while that session is still waiting for a PFCP reply
and has an active PFCP xact.
In this case, xact->data points to the deleted session
and xact's timeout function (sess_5gc_timeout for example)
eventually refers to this already freed session.
This fix prevents duplicate deletes from occurring by checking to see
if the session context has already been deleted when the timeout occurs.
Additionally, it moves session deletions out of timer callbacks into
state machine by reselect_upf().
Due to the way 'ogs_timer_mgr_expire' calls timer callbacks,
one must not stop or expire timers from within a timer callback.
And now one must not remove sessions from within a timer callback.
TS23.007 17.4.1
19A PFCP based restart procedures
After a PFCP entity has restarted, it shall immediately update all local Recovery Time Stamps and shall clear all remote
Recovery Time Stamps. When peer PFCP entities information is available, i.e. when the PFCP Association is still alive,
the restarted PFCP entity shall send its updated Recovery Time Stamps in a Heartbeat Request message to the peer
PFCP entities before initiating any PFCP session signalling.
3GPP TS 29.244 7.2.2.4.2 documents that the peer will set SEID=0 in the
response when we request something for a session not existing at the peer.
If that's the case, we still want to locate the local session which
originated the request, so let's store the local SEID in the xact when
submitting the message, so that we can retrieve the related SEID and
find the session if we receive SEID=0.
The use of the Gy interface (SMF acting as CTF towards an OCS node) is
mandated through configuration file. Default value "enable: auto" will
only make use of it in case a Diameter peer announcing support for the
Credit-Control Application is found.
Upon subscriber session creation, and after auth check over Gx, the SMF
will create a Gy session with the OCS and only after that step the SMF
will accept the session back to the subscriber.
The OCS may then grant some traffic volumes/time and ask to be notified
back with updated measurements.
In order to get the measurements, the SMF relies on PFCP URR configured
to the UPF through Session Repoort Request messages.
When closing the subscriber session, the SMF will also terminate the Gy
session at the OCS.
So far only some specifics parts of the Gy interface as well as the PFCP
side are implemented. Those should be enough to at least have
volume/time thresholds granted by the OCS, which then will be able to
track subsriber resource use.
This patch doesn't implement the OCS side of the Gy interface, that's
left as a future exercise. The interface was tested using an OCS
emulator implemented in TTCN-3 [1]
[1] https://cgit.osmocom.org/osmo-ttcn3-hacks/
TODO:
* Use an event for the report, like SMF_EVT_N4_TIMER?
* Properly set Service identifier in Gy CCR
* SMF: Properly set pkt/octet volumes in Gy CCR
** Update when receiving PFCP Modify Response.
* Figure out best way to require Gy through config file in open5gs-smfd.
* Create a new sess-sm.c which handles smf_sess_t state through
Gx+Gy+PFCP creation, modification and tear down. This way we can do
stuff in parallel, for instance Gx+Gy. It will alsoavoid duplicating
some code paths due to Gy being optional.
So far, no operation was performed when Error Indication was received
from eNodeB. For that reason, I solved #568 issues by controlling
the MME to prevent this from happening.
Now, when GTP-U Error Indication is received, MME and SGW are implemented
to do what they have to do. I hope that the network can be restored
by responding appropriately even if Error Indication occurs.
- Set the number of UEs in units of AMF/MME instead of gNB/eNB.
- See default value as shown below
Number of UEs per AMF/MME : 4,096
Number of gNB/eNB per AMF/MME : 32
Sukchan Lee
Pau Espin