Survey							
                            
		                
		                * Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Requirements for the Resilience of Control Plane in GMPLS (draft-kim-ccamp-cpr-reqts-00.txt) CCAMP WG (59th IETF) Apr.04, 2004 Young Hwa Kim ([email protected]) Contents 1. Summary of the draft 2. Necessities for Resilience of Control Networks 3. Conclusion 59th IETF(CCAMP WG) (2/9) 1. Summary of the draft (2/1) ▣ 3 components of CP : entities, nodes, and channels ▣ Control networks  Control modes  Network configurations ▣ Some concepts for the resilience of control networks  Control packet type  Active and standby control channels  Protection group … ▣ Requirements for the resilience of control networks  Configuration of control networks  Priorities in control channels  Reverting and non-reverting modes of control channels 59th IETF(CCAMP WG) (3/9) 1. Summary of the draft (2/2) ▣ Necessities for the resilience of control networks ▣ Relation to LMP ▣ Possible functions for the resilience of control networks  Identification of an active control channel  Negotiation of switchover attributes  Verification of standby control channels  Automatic switchover  Forced switchover  Inquiry of switchover attributes  Notification of protocol errors  Parameter negotiation and Hello protocol 59th IETF(CCAMP WG) (4/9) 2. Necessities for Resilience of Control Plane (1/3) ▣ Current status of control plane  Weak concept of control network  Slow convergence of IP routing protocols due to hello intervals and expiration timer values  Alternative diverse paths between communication entities  A control network to experience an independent failure from the transport network, and its impact on connection setup and teardown requests  Also, control channels to share common physical routes with the transport network, and its impact on restoration of existing connections  Separation of control channels in GMPLS  Control channel management in LMP 59th IETF(CCAMP WG) (5/9) 2. Necessities for Resilience of Control Plane (2/3) ▣ Control channel management in LMP  Control channel activation with parameter negotiation  Hello protocol  Control channels independent of TE links  Possible multiple active control channels between a pair of nodes  If a failure on an active control channel occurs, alternative active control channels can be used, or the activation procedure can be performed.  No guarantee of the same level of service about TE links in a Degraded state 59th IETF(CCAMP WG) (6/9) 2. Necessities for Resilience of Control Plane (3/3) ▣ Is the status satisfactory? No. Then, why?  Few consideration of control modes and several control network configurations  While admitting the resilience of transport plane, no care of the resilience of control plane  No resilience concept in LMP  If there is no active control channel under the situation that a control channel is responsible for several TE links, the links fall into the Degraded state. We do not want it.  Even when there is a failure of an active control channel on restoration of existing connections, we do not want to start the control channel activation again, and we should be capable of identifying an alternative active control channel promptly. 59th IETF(CCAMP WG) (7/9) 3. Conclusion ▣ The draft introduced a framework for the resilience of control plane. ▣ The control channel management in LMP is not sufficient for the resilience of control plane in GMPLS. ▣ The resilience of transport plane may be included the current charter of the CCAMP WG, but whether the one of control plane is included in the WG or not … ▣ If agreed to this point, let’s put the resilience of control plane into a basket of re-chartering items. ▣ Then, I propose this draft as a WG document of the CCAMP WG. ▣ Future works  Refinement of this draft; and  Proposal of a preliminary protocol specification. 59th IETF(CCAMP WG) (8/9) Thank you. Young Hwa Kim ([email protected])