Enhanced General Switch Management Protocol Salim Hariri Department

Enhanced General Switch Management Protocol Salim Hariri Department of Electrical and Computer Engineering HPDC Laboratory The University of Arizona ECE Building, Room 421 Tucson, Arizona, AZ 85750 Tel: (520) 621 -4378, Fax: (520) 621 -8076 hariri@ece. arizona. edu, www. ece. arizona. edu/~hpdc Salim Hariri HPDC Laboratory http: //www. ece. arizona. edu/~hpdc

Why Programmable Networks? l l l Rapid creation, deployment and management of new services in response to user demands. Change in the nature of traffic due to the wide variety of applications and services. Application specific demands for resources. Need for the separation of communication hardware from control software. Better control over the network resources for its effective use. Salim Hariri HPDC Laboratory http: //www. ece. arizona. edu/~hpdc

Classification of Programmable Networks Open Interface Networks (Static Approach) ATM GSMP Salim Hariri q-GSMP IP P 1520 e-GSMP Model HPDC Laboratory Active Networks (Dynamic Approach) Discrete Approach (Out of Band) Integrated Approach (In band) http: //www. ece. arizona. edu/~hpdc

Open Interface Networks l Provides abstractions in the layers of a node to define programmable interfaces. l Allows applications and middle-ware to manipulate low-level network resources. l Uses APIs to control the various layers. Salim Hariri HPDC Laboratory http: //www. ece. arizona. edu/~hpdc

Open Interface Networks (Contd. ) l Advantages: – Separation of service business. – Separation of vendor business. – Faster standardization. – Extensibility – Richer Semantics Salim Hariri HPDC Laboratory http: //www. ece. arizona. edu/~hpdc

Meaning of Qo. S l Quality of Service (Qo. S) refers to the capability of the network to provide better service to selected network traffic irrespective of the underlying technologies. l The goal of Qo. S is to provide priority including dedicated bandwidth, controlled jitter and latency and improved loss characteristics. Salim Hariri HPDC Laboratory http: //www. ece. arizona. edu/~hpdc

Qo. S in OSI Model Salim Hariri HPDC Laboratory http: //www. ece. arizona. edu/~hpdc

Qo. S over ATM l Providing Qo. S over ATM is easier – Fixed length of cells – Well defined types of services. l Techniques for providing Qo. S in ATM – General Switch Management Protocol (GSMP) – q-GSMP Salim Hariri HPDC Laboratory http: //www. ece. arizona. edu/~hpdc

General Switch Management Protocol(RFC 1987) l l l Open Interface, switch control protocol. Connection oriented network technologies. Point-point and Multi-point connections. Adjacency protocol -synchronize state across link. Master-Slave relationship between controller and the switches Salim Hariri HPDC Laboratory http: //www. ece. arizona. edu/~hpdc

GSMP VPC/VCCs Policer Classifier Regulator Scheduler Confirming Qo. S Class Excess Output Port Confirming Qo. S Class Salim Hariri HPDC Laboratory Excess http: //www. ece. arizona. edu/~hpdc

GSMP Message Sequence Connection Management Port Management Configuration Information Switch controller Statistics Events Network q. GSMP Salim Hariri Switch Mapping HPDC Laboratory ATM Switch http: //www. ece. arizona. edu/~hpdc

q-GSMP l l l Provides Qo. S extensions to the GSMP. Supports new messages enabling selection of: – Qo. S constraints – Buffer management – Scheduling algorithms – Memory allocation schemes. Specific to ATM switches. Salim Hariri HPDC Laboratory http: //www. ece. arizona. edu/~hpdc

q-GSMP Connection Management Port Management Configuration Information Switch Controller Statistics Events Qo. S Management Qo. S Configuration Management Qo. S Statistics Qo. S Events q. GSMP Salim Hariri HPDC Laboratory Switch Mapping Scheduler Network Buffer Manager SR Estimator ATM Switch http: //www. ece. arizona. edu/~hpdc

Qo. S over IP Issues: l IP was not meant to provide better than the best effort Qo. S. l Complex scheduling and buffer management due to variable length packets. l Changes in the traffic pattern at the output of a router due to traffic aggregation. Salim Hariri HPDC Laboratory http: //www. ece. arizona. edu/~hpdc

Need for IP Qo. S l Traditional IP networks support only best effort delivery. l Convergence of voice and data requires IP network to provide deterministic guarantees for real time traffic. l Multimedia traffic require both bandwidth and delay guarantees. Salim Hariri HPDC Laboratory http: //www. ece. arizona. edu/~hpdc

P-1520 Model l l Defines a set of Programmable interfaces for the development of the protocol and management of the infrastructure. Defines four interfaces: – Value Added Service Level – Network Generic Services Level – Virtual Network Device Level – Physical Elements Level Salim Hariri HPDC Laboratory http: //www. ece. arizona. edu/~hpdc

P-1520 Model V Interface Algorithms for value added communications U Interface L Interface CCM Interface Algorithms for routing and connection management Network Generic Services Level Virtual Network Devices Level Physical Elements (Hardware) Salim Hariri Value Added Services Level HPDC Laboratory PE Level http: //www. ece. arizona. edu/~hpdc

IP reference Model Salim Hariri HPDC Laboratory http: //www. ece. arizona. edu/~hpdc

Features of our IP Model l Extension of Integrated Services Model of Clark, Shenker and Zhang[ ]. l Conforms to P 1520 model[ ] of programmability. l Complies with For. CES architecture[ ] of network entity. l Incorporates ETRI’s Router Architecture. Salim Hariri HPDC Laboratory http: //www. ece. arizona. edu/~hpdc

ETRI’s Router Architecture Salim Hariri HPDC Laboratory http: //www. ece. arizona. edu/~hpdc

Scheduler and Queues Scheduler Outbound Packets in Various Queues Salim Hariri HPDC Laboratory http: //www. ece. arizona. edu/~hpdc

Algorithm for Schedulability We consider the Diffserv Qo. S, classes of service. Let us assume that DS(i) = ith class of service. Vector Q(i) = [ q 1(i) q 2(i), ……. qm(I)] be the Qo. S Parameters for the ith class. We define a matrix Qm consisting of all the Qo. S classes and their respective parameters. Q(1) Qm = Q(2) Q(n) Salim Hariri q 1(1) q 2(1) ………qm(1) = q 1(2) q 2(2) ………qm(2) q 1(n) q 2(n) ………qm(n) HPDC Laboratory http: //www. ece. arizona. edu/~hpdc

Input of a Router 01 02 Media Packet I/F Classification Inbound Packets Switch/ Backplane n-1 n Bandwidth/Delay Shaped Queues Salim Hariri HPDC Laboratory Scheduler Per-dest Queues http: //www. ece. arizona. edu/~hpdc

A table K is maintained by the router to indicate the number of packets in each Qo. S class. K = [k 1, k 2…………kn] Bo = port bandwidth B = total bandwidth. Bo = Σ ki * q 1(i) if (q 1(i) <= (B-Bo)) then Band. Schedulable =TRUE; else Band. Schedulable = FALSE; Salim Hariri HPDC Laboratory http: //www. ece. arizona. edu/~hpdc

Output from a Router Packet Media Processor I/F Outbound Packets Switch/ Backplane Bandwidth/Delay Shaped Queues Salim Hariri HPDC Laboratory Scheduler http: //www. ece. arizona. edu/~hpdc

Admission Control Algorithm The Admission control is made at the line cards. Token Bucket algorithm is used to characterize the flows. A new flow is admitted if the following condition is satisfied PBRnew + NΣ i=1 PBRi <= ρC where PBR = peak bit rate ρ = admissible load of capacity C = total capacity Salim Hariri HPDC Laboratory http: //www. ece. arizona. edu/~hpdc

e-GSMP l l Envisages to provide Qo. S services over a primarily IP network. Is an Open Interface approach. Defines a Master-Slave relationship between the Controller and the IP Routers. Allows an interactive approach to provide programmability. Salim Hariri HPDC Laboratory http: //www. ece. arizona. edu/~hpdc

e-GSMP Switch Mapping 1 Schedulable Region Estimators Multiplexer m Input Ports Salim Hariri Multiplexer Switch Fabric HPDC Laboratory Output Ports http: //www. ece. arizona. edu/~hpdc

e-GSMP l l l To support the Intserv and Diffserv frameworks. To define schedulable regions to implement admission control. To allow for different scheduling and buffer management techniques. Salim Hariri HPDC Laboratory http: //www. ece. arizona. edu/~hpdc

Concluding Remarks Salim Hariri HPDC Laboratory http: //www. ece. arizona. edu/~hpdc