для Cisco Cisco Packet Data Gateway (PDG)
P-GW Changes in Release 16
P-GW Enhancements for 16.0 ▀
Release Change Reference, StarOS Release 16 ▄
363
The following output had been added to show GTP-U source port configuration.
GTPU Source-port Configuration
CSCuh35139 - PGW SM sync up with driver [Phase 1]
Feature Changes
Change in Message Generation Toward Access Side
Previous Behavior:
1. In case of GTP-U error indication, P-GW was sending out Delete Bearer Request and was waiting for response
before doing the clean up.
2. If there was a collision between GTP messages and the older transaction got aborted, until now any PCRF rules
coming in the response for that transaction were getting aborted.
New Behavior:
1. In case of GTP-U error indication, P-GW will align with GGSN and not send any Delete Bearer Request out. It
will do the required clean up locally.
2. If there is a collision between GTP messages and the older transaction gets aborted, the old behavior will be
retained if the transaction was pending at Gy/Rf when it got aborted. If, however, it was pending at PCRF,
these rules will now be processed.
these rules will now be processed.
Customer Impact:
1. In case of GTP-U error indication, the message was not needed since the peer does not have that particular
bearer.
2. Customer will start seeing some rules processed, in case of collision, which were not processed before.
CSCuh85489, CSCui06202 - SR Performance: Periodic FC interval needs to be
increased.
Feature Changes
Increased Periodic Full Checkpoint (FC) interval
The Background Periodic Full Checkpoint (FC) interval (automatically triggered by the system), has been increased to
15 minutes from the earlier value of 12 minutes. This was done in order to save some CPU cycles. The applications send
FC when there is a change in the status of the UE and applications also update intermediate status by sending micro
checkpoints. Therefore, the background periodic full checkpoint interval has been increased.
15 minutes from the earlier value of 12 minutes. This was done in order to save some CPU cycles. The applications send
FC when there is a change in the status of the UE and applications also update intermediate status by sending micro
checkpoints. Therefore, the background periodic full checkpoint interval has been increased.
For SR (session recovery) case, this change implies that the system automatically triggers the full check pointing to
AAA Manager every 15 minutes.
AAA Manager every 15 minutes.
For ICSR case, this change implies that the system automatically triggers Full check pointing to the standby chassis
every 45 minutes instead of the earlier 36 minutes.
every 45 minutes instead of the earlier 36 minutes.
Important:
This time interval is not configurable.