Content for TR 29.816 Word version: 10.0.0
2 References
3 Definitions and abbreviations
4 Failure and restoration scenarios
5 Functional requirements for solutions
6 Solutions
7 Evaluation
8 Conclusion
A coding examples
$ Change history
2 References p. 8
3 Definitions and abbreviations p. 9
3.1 Definitions p. 9
3.2 Abbreviations p. 9
4 Failure and restoration scenarios p. 10
4.1 Baseline architecture p. 10
4.2 Failure and recovery detection p. 13
4.2.1 General p. 13
4.2.2 Failure detection on transport level p. 14
4.2.3 Failure and recovery detection on DIAMETER level p. 14
4.2.4 Failure and recovery detection on PCRF application level p. 15
4.3 PCRF node failure scenarios p. 15
4.4 Selection of restoration scheme p. 18
5 Functional requirements for solutions p. 19
5.1 General p. 19
5.2 Functional requirements for all deployments p. 19
5.3 Functional requirements for the multiple PCRF deployment with DRA p. 19
6 Solutions p. 20
6.0 General p. 20
6.1 Solution 1: Solution for the PCRF failure reselection for the DRA p. 20
6.1.1 Redirect DRA p. 20
6.1.2 Proxy DRA p. 21
6.2 Solution 2: use of DIAMETER base protocol in Single PCRF deployment (with Direct Client-Server Connection) p. 23
6.2.1 PCRF Failure Detection p. 23
6.2.2 PCRF Restart p. 23
6.3 Solution 3: Graceful termination of services p. 23
6.3.1 General p. 23
6.3.2 Graceful termination in PCEF p. 23
6.3.3 Graceful termination in BBERF (Serving GW) p. 24
6.3.4 Graceful termination in AF p. 24
6.4 Solution 4: Strict termination of bearer services p. 24
6.5 Solution 5: PCRF session state restoration p. 25
6.6 Solution 6: Soft recovery after a PCRF restart p. 28
6.6.1 Role of PCRF p. 28
6.6.2 Actions required for a soft recovery p. 29
6.6.2.1 Procedure p. 29
6.6.2.2 Binding p. 30
6.6.2.2.1 Status after restart p. 30
6.6.2.2.2 OPTION 1: Identities exchange with recovery/rebuild messages p. 30
6.6.2.2.3 OPTION 2: Restore/rebuild request sent to all candidates p. 31
6.6.2.2.4 OPTION 3: Related Diameter identities saved and retrieved p. 32
6.6.2.3 Information exchange between PCRF and clients p. 33
6.6.2.3.1 Status after restart p. 33
6.6.2.3.2 Information from AF p. 33
6.6.2.3.3 Information from PCEF p. 33
6.6.2.3.3.1 OPTION 1: PCRF retrieves input parameters and recreates the lost information p. 34
6.6.2.3.3.2 OPTION 2: PCRF retrieves information sent to PCEF before restart p. 34
6.6.2.3.4 Information from BBERF p. 34
6.6.2.3.5 Information from PCRF p. 34
6.6.2.3.5.1 Restarted V-PCRF, home routed access p. 35
6.6.2.3.5.2 Restarted V-PCRF, visited access p. 35
6.6.2.3.5.3 Restarted H-PCRF, home routed access p. 35
6.6.2.3.5.4 Restarted H-PCRF, visited access, AF in HPLMN p. 35
6.6.2.3.5.5 Restarted H-PCRF, visited access, AF in VPLMN p. 35
6.6.2.3.6 Information from SPR p. 35
6.6.2.3.7 Messages for information transfer p. 35
6.6.3 Impact of the solution on specifications p. 35
6.7 Solution 7: Bulk Signaling p. 37
6.7.1 General bulk signaling p. 37
6.7.2 Bulk signalling based on PCRF Session Set ID (PSSID) p. 37
6.7.2.1 Concept p. 37
6.7.2.2 Use in signaling p. 38
6.7.3 Embedding in the DIAMETER signaling concept p. 39
6.8 Solution 8: Adding explicit resilience to PCRF sessions p. 39
6.8.1 Concept p. 39
6.8.2 Signaling procedures p. 40
6.9 Solution 9: Unified solution for termination of bearer services p. 43
7 Evaluation p. 46
7.1 General p. 46
7.2 Comparison of restoration solutions p. 47
7.3 Evaluation of failure and recovery detection mechanisms p. 49
8 Conclusion p. 51
A coding examples p. 52
A.1 Bulk signaling p. 52
A.2 Restart indication p. 54
A.3 PCRF session state restoration p. 58
A.4 Adding explicit resilience to PCRF sessions p. 60
$ Change history p. 64
