Volker Hilt Bell Labs/Alcatel-Lucent SIP Overload Control IETF Design Team Status

Slide 2 | 74 IETF Meeting | March 2009 SIP Overload Control Design Team Team Members  Eric Noel, Carolyn Johnson (AT&T Labs)  Volker Hilt, Fangzhe Chang (Bell Labs/Alcatel-Lucent)  Charles Shen, Henning Schulzrinne (Columbia University)  Ahmed Abdelal, Tom Phelan (Sonus Networks)  Mary Barnes (Nortel)  Jonathan Rosenberg (Cisco)  Nick Stewart (British Telecom) Four independent simulation tools  AT&T Labs, Bell Labs/Alcatel-Lucent, Columbia University, Sonus Networks Bi-weekly conference calls.

Slide 3 | 74 IETF Meeting | March 2009 draft-ietf-sipping-overload-design-01 Changes to -00 Added new sections on:  Fairness  Introduces fairness categories.  Performance Metrics  Discusses metrics to compare overload control mechanisms  Message Priorization  Selection of messages in overload condition. Added text to Security Considerations section. Minor edits throughout the text.

Slide 4 | 74 IETF Meeting | March 2009 draft-ietf-sipping-overload-design-00 Next Steps Discussion of overload control mechanisms needs to be structured along the identified performance metrics. Document is close to completion.

Slide 5 | 74 IETF Meeting | March 2009 SIP Overload Control Design Team Simulation Results Four types of overload control  Rate-based Overload Control  Loss-based Overload Control  Window-based Overload Control  Overload Signal-based Overload Control Summary of Steady-State Evaluation (presented at IETF ’73) Performance of all overload control mechanisms under evaluation is similar in steady state. Varying network conditions (i.e., delay, loss-rate) do not reveal significant differences. Results for Transient Scenarios Evaluation of transient behavior with respect to  Changes in offered load  changes in the number of neighbors  Fairness

Slide 6 | 74 IETF Meeting | March 2009 Changes in Offered-Load (AT&T Labs) Rate-based and Window-based Overload Control Simulations use the following overload control feedback types and algorithms:  Rate-based: queue delay  Loss-based: SRED  Window-based Feedback conveyed in SIP responses. Result: rate-, loss- and window-based controls respond well to transient stimulus.

Slide 7 | 74 IETF Meeting | March 2009 Changes in Offered-Load (Bell Labs/Alcatel-Lucent) Loss-based and Rate-based Overload Control Time Overload control feedback type and algorithms used:  SRED algorithm  Loss- vs. rate-based feedback SIP responses convey feedback from core to edge proxies. Result: loss- and rate-based overload control perform well. Time CPS

Slide 8 | 74 IETF Meeting | March 2009 Changes in Offered-Load (Columbia University) Window-based and Rate-based Overload Control Slide 8  Window- and rate-based controls perform well.

Slide 9 | 74 IETF Meeting | March 2009 Changes in Offered-Load (Sonus Networks) Loss-based and Overload-Signal-based Overload Control Target Overload Signal Rate =10 Overload Signals /Sec

Slide 10 | 74 IETF Meeting | March 2009 Changes in the Number of Senders (Bell Labs/Alcatel-Lucent) Loss-based and Rate-based Overload Control Edge proxies are turned on/off sequentially.  Each edge proxy sends the same amount of load while active. Feedback-type and algorithms:  Rate-fixed: core proxies are configured with a fixed number of senders. The overall rate of a core proxy is divided through the sender number.  Rate-aware: core proxies estimate the number of senders. The overall rate of a core proxy is divided through the sender estimate.  Loss-based: same loss rate is sent to all edge proxies.  All simulations use SRED algorithm.

Slide 11 | 74 IETF Meeting | March 2009 Fairness (Columbia University) Rate-based Overload Control Slide 11  Provider-centric fairness: each source gets the same share  User-centric fairness: each source gets a share proportional to its original incoming load

Slide 12 | 74 IETF Meeting | March 2009 Conclusion & Next Steps Simulation Results The overload control performance seems to differ little between the type of feedback:  Rate-, Loss-, Window- and Signal-based mechanisms all performed well in steady-state as well as transient evaluations. Of course, the performance does vary depending on the overload control algorithms used and parameter settings of these algorithms.  But: algorithms and parameter settings are likely to be out of scope for an overload control protocol specification. Next Steps Evaluate additional transient scenarios. Finalize draft-ietf-sipping-overload-design-01 Work on a solution!!

Session Initiation Protocol (SIP) Overload Control draft-hilt-sipping-overload-06 Volker Hilt, Indra Widjaja, Henning Schulzrinne A Session Initiation Protocol (SIP) Load Control Event Package draft-shen-sipping-load-control-event-package-01 Charles Shen, Henning Schulzrinne, Arata Koike

Slide 14 | 74 IETF Meeting | March Service Provider B Hotline Callee am-10am, Service Provider A Enterprise Network B Enterprise Network A Filter Spec ID: To: Time: 9am-10am Act: accept rate= N max Filter Spec ID: To: Time: 9am-10am Act: accept rate=N SPA Filter Spec ID: To: Time: 9am-10am Act: accept rate=N EPA Filter Spec ID: To: Time: 9am-10am Act: accept rate=N SPB Charles Shen, Henning Schulzrinne, Arata Koike, A Session Initiation Protocol (SIP) Load Control Event Package, draft-shen-sipping-load-control-event-package-01.txt, IETF SIPPING Working Group, Work in Progress. Nov 3, 2008 Filter-based SIP Server Overload Control draft-shen-sipping-load-control-event-package-01

Slide 15 | 74 IETF Meeting | March 2009 Server S1Server S2 Session Initiation Protocol (SIP) Overload Control draft-hilt-sipping-overload-06 Overload control mechanism  Enables proxies to send overload control feedback to upstream neighbors.  Feedback is conveyed in SIP responses  New Via Header Parameters  Supports different types of feedback. Currently defined: loss-based.  Specifies the protocol semantics. Open to different overload control algorithms. Via: SIP/2.0/TCP ss1.example.com:5060 ;oc=20;oc_validity=500 OverloadReduces load