Sun Ethernet Fabric Operating System PVRST+ Administration Guide

Exit Print View

Updated: July 2015
 
 

Rapid STP

The bridge allows interconnection of end stations attached to separate LANs and allows the stations to communicate as if they are attached to a single LAN. The bridge operates below the MAC service boundary and is transparent to the protocols operating above this boundary.

In complex networks, a loop can occur when there are two or more paths between two end points. This situation leads to the duplication of frames, which in turn leads to heavy traffic in the network. To avoid this situation, STP is used. STP forms a logical, loop-free topology from the physical topology and forwards the frames without duplication. To avoid prolonged stabilization time following a reconfiguration event in the spanning tree algorithm, PVRST+ provides support for RSTP. The operation of RSTP enables rapid recovery of connectivity following the failure of a bridge, bridge port, or a LAN.

To isolate link fluctuations to a particular VLAN segment and to provide load balancing, PVRST+ creates a separate spanning tree for each VLAN. The spanning-tree-to-VLAN mapping is configured on a per-VLAN basis.

The PVRST+ protocol is an enhancement of RSTP, which works in conjunction with VLAN to provide better control over traffic in the network.

A switch takes the role of either a root or designated switch. STP assigns port roles to the port to calculate the best loop-free path. This list contains the possible roles:

  • Root - Port that offers the lowest cost path towards the root bridge.

  • Designated - Forwarding port elected for every switched LAN segment.

  • Alternate - Blocked port providing an alternate path to the root bridge of the spanning tree.

  • Backup - Blocked port that acts as a backup for the path provided by a designated port.

These elements determine the stable and active spanning-tree topology of a switched network.

  • Bridge ID (instance ID, switch priority, and MAC address)

  • Path cost to the root switch maintained for each instance

  • Port identifier (port priority and the port number)

  • Designated bridge ID (the bridge through which the root bridge is connected)

When switches in a network come up, each switch assumes itself to be the root bridge and starts sending configuration messages through all of its ports. BPDUs are used to communicate and compute the spanning tree topology. The BPDUs contain the following information:

  • VLAN ID to which the BPDU belongs

  • Unique bridge ID of the switch that has been identified as the root

  • Spanning-tree path cost to the root

  • Bridge ID of the sending switch

  • Message age

  • Identifier of the sending interface (port priority and port number)

  • Values for the hello, forward-delay, and max-age protocol timers

When a switch receives a superior configuration BPDU on a port, it stores the received information for that port. If the port is a root port, the switch forwards the updated message to all of the attached LANs for which the switch is the designated bridge. If the switch receives an inferior configuration BPDU to the BPDU that is currently stored for that port, the switch discards the BPDU. If the switch is a designated switch for the LAN from which the inferior information was received, then the switch sends up-to-date information stored for that port, discards the inferior information, and propagates the superior information in the network.

Each layer 2 interface in the switch running the per-VLAN rapid spanning tree protocol is in one of these states for each VLAN running on the switch:

  • Blocking - The interface discards the frames and does not learn the MAC addresses.

  • Listening - This is the first state in which a port transit resides after the port has been blocked. The interface enters this state when STP decides that the interface must participate in frame forwarding.

  • Learning - An interface enters this state from the listening state. The interface gets ready to participate in frame forwarding and learns MAC addresses from the received packets.

  • Forwarding - The interface receives frames on the port and forwards the frames switched from another port. The transition from blocking to forwarding takes 30 seconds.

These sections describe switch priorities and election, port states and roles, rapid convergence, proposal agreement and how to detect topology change.