Implementation and simulation of Stacking Protocol in OPNET

1
Implementation and simulation of Stacking
Protocol in OPNET

• Project by Itamar Cohen
• Supervisor Nir Arad

2
Project General Description

• Prestera is novel high-performance chipset
  architecture by MARVELL.
• Prestera Stacking Protocol (SP) is used to
  configure several switching devices as a single
  network entity.
• SP operates on top of two transport protocols,
  named RMF and FMF.
• In this project RMF and FMF will be implemented
  and tested in the OPNET Modeler.

3
RMF (Reliable Message Forwarding) protocol

• Responsible for hop-to-hop communications between
  switching devices.
• based on Stop Wait algorithm.
• Handles SAR (segmentation and reassembly) of
  packets.
• Has interfaces to the sending application and to
  the hardware driver, which receives the packets
  from the stream.

4
S W demonstration
5
S W demonstration (Cont)
6
(No Transcript)
7
(No Transcript)
8
RMF implementation issues

• A single RMF process serves a few application and
  uses a few links.
• Each link has its own parameters
• Number of retries
• I.D. of the requested cell (0 or 1).
• Identity of the destination application.
• Each link and each application has its own SAR
  buffer.
• Time out counting is executed by scheduling
  future events, and canceling them if the ACK has
  already arrived.

9
FMF (Flood Message Forwarding) protocol

• Built above the RMF.
• Based on the PI (Propagation of Information) and
  PIF (Propagation of Information with Feedback)
  algorithms.
• Handles the whole-network requirements.

10
PI - definitions

• MSG (info) a message carrying the information
  info to be propagated.
• Gi - set of neighbors of node i.
• mi indicates whether node i has already entered
  the protocol.

11
IP the protocol for every node

• Init for every node i mi 0
• While receiving a MSG (info)
• If (mi 0)
• mi 1
• Accept (info)
• For each k in Gi send MSG (info) to k

12
SP (Stacking Protocol)

• Is responsible for the following tasks
• Studying the network topology.
• Configure the units in the network according to
  the topology.
• Handle dynamic changes in the topology.