• Composante

    École Nationale Supérieure d'Électrotechnique d'Électronique d'Informatique d'Hydraulique et des Télécommunications

Objectifs

  • To present new paradigm of networking : virtualisation control and  programmability
  •  To detail new technologies for control (QoS ,Routing, Reliability)  DCBx, Trill, segment routing, EPR
  • To explain cloud networking architectures
Lire plus

Description

Chap1 Context-

Data Center history: data and storage networks-from centralised to cloud networking Virtualisation  principle of containers versus virtual machine-levels of virtual communications-specificities of L 2  virtual communications MAC addressing and extended VLAN segmentation  L3 virtual communications addressing, IP floating .

Chap2 Data Center Architecture

DC Network. Network element architecture- Isolated processes: routing and switching- Integrated processes: flow switching- Separated processes : hardware design with multistage Fabric and software design with SDN concept

 Performance of communication aArchitecture-Limitations- Congestion management principles- Traffic management principles

Chap3 Data center Communication

 Topology types: big switch, clos network fat tree- Routing and topology-Problems and new solutions for DC-Hierarchical L2Routing with Pod and pseudo addresses- L3 routing on pseudo IP addresses

Infrastructure Standardisation TIA942 standard elements-Redundancy and reliability levels-Rated (tiers) DC --Data Center Bridging  standards- Enhanced Ethernet flow control and congestion management. PFC,ETS,CN,DCBX

Chap4 Load sharing

Principles of load sharing: Load sharing objective, levels and processes: discovery, distribution, type of distribution: traffic independent, traffic dependant, load dependant

Path load sharing: Bridging sharing, STP, Trill SPB Packet; Routing sharing- ECMP, Flow routing sharing : MPLS-TE and Segment routing

Chap5 Reliability

Principles : Failure characteristics, fault handling strategies, protection and restoration

Failure recovery for routed network: recovery methods in MPLS-TE, recovery cycles, local versus global strategies, bandwidth sharing versus protection ; Fast reroute Mechanism – overview , types of protection; detour and bypass illustration,  signalling

Failure recovery for bridged  networks: EPRS

Lire plus

Pré-requis obligatoires

  • LAN networking : Ethernet, switching
  • Internet networking IP, TCP,BGP, MPLS
Lire plus

Informations complémentaires