Understanding VBST
VBST is equivalent to running STP or RSTP in each VLAN so that spanning trees in different VLANs are independent of each other. Though VBST does not provide multi-instance, VBST implements load balancing of traffic from different VLANs.
- One root bridge
- Two measurements: ID and path cost
- Three port statuses: Discarding, Learning, and Forwarding
- Five port roles: root port, alternate port, backup port, designated port, and edge port
- Three timers: Hello Time, Forward Delay, and Max Age
-
In VBST, the BID consists of the bridge priority, VLAN ID, and bridge MAC address. The bridge priority occupies the most significant 4 bits, the VLAN ID occupies the 12 bits following the bridge priority, and the MAC address occupies the least significant 48 bits.
On a VBST network, the device with the smallest bridge ID will be selected as the root bridge.
- VBST transmits VBST BPDUs in VLANs to determine the network topology. VBST BPDUs are based on STP/RSTP BPDUs and a 4-byte 802.1q tag is added between the source MAC address and protocol length. Figure 11-1 shows the comparisons between the STP/RSTP BPDU and VBST BPDU.
The DMAC identifies the destination MAC address of packets. The DMAC in a VBST BPDU is 0100-0CCC-CCCD; the Data field in a standard RSTP/STP BPDU is used as the Data field in a VBST BPDU. By default, the Data field in a standard RSTP BPDU is used as the Data field in a VBST BPDU.
VBST implements VLAN-based spanning tree calculation, topology convergence, and interworking with spanning tree protocols of other vendors.
VBST Topology Calculation
VBST supports VLAN-based topology calculation. Tagged VBST BPDUs are sent in each VLAN except VLAN1 and topology calculation is performed separately. The VBST topology calculation method is similar to the STP/RSTP calculation method. For details, see STP Topology Calculation. Different root bridges can be selected in VLANs. Figure 11-2 shows the topology calculation results of STP/RSTP and VBST.
Through topology calculation, STP/RSTP generates a spanning tree with the root bridge as SwitchF. The links between SwitchB and SwitchE and between SwitchA and SwitchD are blocked. HostA and HostB belong to VLAN2. The link between SwitchB and SwitchE does not permit packets of VLAN2 to pass through because the link between SwitchB and SwitchE is blocked. Therefore, HostA fails to communicate with HostB.
Through topology calculation, VBST generates spanning trees VLAN2 and VLAN3 with root bridges as SwitchD and SwitchF respectively. Traffic in VLAN2 and VLAN3 is forwarded through their respective spanning trees so that traffic is load balanced between paths SwitchB-SwitchE and SwitchC-SwitchF.
Fast Convergence of VBST
Common mode
The Proposal/Agreement mechanism in common mode supported by VBST is similar to that supported by RSTP. For details, see Technical Details of RSTP.
Enhanced mode
The Proposal/Agreement mechanism in enhanced mode supported by VBST is similar to that supported by MSTP. For details, see MSTP Fast Convergence.
Protection Mechanisms of VBST
Similar to RSTP, VBST provides BPDU protection, TC protection, root protection, and loop protection. For details, see Protection functions.
Interworking Between VBST and Standard STP/RSTP
On a live network, VBST-enabled devices may connect to STP/RSTP-enabled devices. VBST and STP/RSTP use different BPDU formats, so there are interworking problems. To implement interworking between VBST and standard STP/RSTP, take the following measures:
On a trunk interface:
- When a VBST-enabled device connects to an RSTP-enabled device, the VBST-enabled device uses standard RSTP BPDUs in VLAN1 and VBST BPDUs with the Data field of RSTP BPDUs in other VLANs to exchange with the RSTP-enabled device.
- When a VBST-enabled device connects to an STP-enabled device, the VBST-enabled device uses standard STP BPDUs in VLAN1 and VBST BPDUs with the Data field of STP BPDUs in other VLANs to exchange with the STP-enabled device.
The following describes spanning tree implementation, as shown in Figure 11-3.
As shown in Figure 11-3, STP/RSTP is deployed on SwitchA and SwitchB, and VBST is deployed on SwitchC and SwitchD. Devices are connected through trunk interfaces, and interfaces on SwitchA through SwitchD allow packets from VLAN1 and VLAN10 to pass through.
An STP/RSTP-enabled device can only send and receive STP/RSTP BPDUs, and transparently transmit VBST BPDUs, so a spanning tree is formed in VLAN1 as defined by STP/RSTP.
Assume that the congestion point of the spanning tree in VLAN1 is on SwitchD. Because VBST runs on SwitchD, so the congestion point exists in VLAN1. SwitchD can still receive and forward VBST BPDUs in VLAN10. Loops occur in VLAN10, so spanning tree calculation in VLAN10 is triggered. SwitchA and SwitchB transparently transmit VBST BPDUs in VLAN10, so only four interfaces on SwitchC and SwitchD participate in spanning tree calculation in VLAN10. Then the spanning trees in VLAN1 and VLAN10 are formed, as shown in Figure 11-3.
Assume that the blocking point of the spanning tree in VLAN1 is on SwitchB. STP/RSTP runs on SwitchB, so the blocking port exists on SwitchB. SwitchB cannot forward VBST BPDUs from VLAN10 and no loop occurs in VLAN10, so spanning tree calculation in VLAN10 is not triggered. VBST BPDUs from VLAN10 can be forwarded along the spanning tree in VLAN1, that is, VLAN10 and VLAN1 share the spanning tree, as shown in Figure 11-3.
On an access interface, a VBST-enabled device uses standard STP or RSTP BPDUs to exchange with the remote end according to the VLAN that the access interface belongs to. Topology calculation is performed as defined by STP/RSTP. Because STP/RSTP does not differentiate VLANs, a spanning tree shared by VLANs is formed.
When a VBST-enabled device connects to an STP/RSTP-enabled device, the trunk interface must be used to connect the two devices and the blocking point must be located on the VBST-enabled device to implement load balancing.
Interworking Between VBST and PVST/PVST+/Rapid PVST+
On a live network, a VBST-enabled device may connect to a device enabled with PVST/PVST+/Rapid PVST+.
Trunk interface
When a VBST-enabled device connects to a device enabled with Rapid PVST+, the VBST-enabled device sends standard RSTP BPDUs (or VBST BPDUs with the Data field of RSTP BPDUs) and VBST BPDUs with the Data field of RSTP BPDUs in other VLANs to exchange with the device enabled with Rapid PVST+.
When a VBST-enabled device connects to a device enabled with PVST+, the VBST-enabled device sends standard STP BPDUs (or VBST BPDUs with the Data field of STP BPDUs) and VBST BPDUs with the Data field of STP BPDUs in other VLANs to exchange with the device enabled with PVST+.
When a VBST-enabled device connects to a PVST-enabled device, packet exchange is similar to that in the scenario where a VBST-enabled device connects to a device enabled with PVST+. The difference is that the VBST-enabled device and PVST-enabled device send only VBST BPDUs with the Data field of STP BPDUs in VLAN1.
The two devices can identify the BPDUs carrying VLAN information, so a VLAN-based spanning tree is formed. The connection between a VBST-enabled device and a device enabled with PVST/PVST+/Rapid PVST+ through a trunk interface is similar to the connection between two VBST-enabled devices.
Access interface
A VBST-enabled device uses standard STP BPDUs to exchange with the device enabled with PVST/PVST+ or RSTP BPDUs to exchange with the device enabled with Rapid PVST+ according to the VLAN that the access interface belongs to. Topology calculation is performed as defined by STP/RSTP. Because STP/RSTP does not differentiate VLANs, a spanning tree shared by VLANs is formed.