안녕하세요.
항상 좋은 자료에 감사드립니다.
GTP End Marker 관련해서 문의 사항이 있어 여러분들의 소중한 지식을 요청드리고자 글을 올립니다.
X2HO나 S1HO시에 만약 End Marker가 누락된다면 어떤 effect가 있는지 문의드립니다. (Call drop??)
End Marker는 제가 이해하기로는 단지 noti가 주 목적이고 이 정보를 통해 Target에서 어디부터 데이터를 수신하고 송신한지를 결정하는데 도움이 되는 것으로 알고 있는대요.
없다면 어디부터 보내고 어디부터 수신할지 모르기에 그냥 Target eNB 입장에서는 buffer에 있는거 그대로 전달하면 될듯 한대요.
혹 자세히 정보를 알려 주실 수 있다면 감사하겠습니다.
10.1.2.2 Path Switch
After the downlink path is switched at the Serving GW downlink packets on the forwarding path and on the new direct path may arrive interchanged at the target eNB. The target eNodeB should first deliver all forwarded packets to the UE before delivering any of the packets received on the new direct path. The method employed in the target eNB to enforce the correct delivery order of packets is outside the scope of the standard.
In order to assist the reordering function in the target eNB, the Serving GW shall send one or more "end marker" packets on the old path immediately after switching the path for each E-RAB of the UE. The "end marker" packet shall not contain user data. The "end marker" is indicated in the GTP header. After completing the sending of the tagged packets the GW shall not send any further user data packets via the old path.
Upon receiving the "end marker" packets, the source eNB shall, if forwarding is activated for that bearer, forward the packet toward the target eNB.
On detection of an "end marker" the target eNB shall discard the end marker packet and initiate any necessary processing to maintain in sequence delivery of user data forwarded over X2 interface and user data received from the serving GW over S1 as a result of the path switch.
On detection of the "end marker", the target eNB may also initiate the release of the data forwarding resource. However, the release of the data forwarding resource is implementation dependent and could also be based on other mechanisms (e.g. timer-based mechanism).
EPC may change the uplink end-point of the tunnels with Path Switch procedure. However, the EPC should keep the old GTP tunnel end-point(s) sufficiently long time in order to minimise the probability of packet losses and avoid unintentional release of respective E-RAB(s).
위 부분을 보시면, 포워딩 패킷 순서 유지는 그냥 out of scope 입니다. 그리고, S-GW에서 트래픽을 절체하면서 End marker를 보내게 되어 있는데, 실지로 유선망의 패킷 로스 확률을 고려해보면, new path로 데이터가 도착하고, old path의 데이터가 안오기 시작하면, Target에서는 일정 딜레이를 주고, new path 데이터를 전송 시작하는 것이 맞아 보입니다.
It is helpful to me.
Thanks.