46
a device that knows nothing about STP or RSTP. Usually, the connected device is an end station. Edge
Ports will immediately transit to forwarding state and skip the listening and learning state because the
edge ports cannot create bridging loops in the network. This will expedite the convergence. When the link
on the edge port toggles, the STP topology remains unchanged. Unlike the designated port or root port
though, an edge port will transit to a normal spanning-tree port immediately if it receives a BPDU. Default
is No.
Admin Point To Point:
A port is a point-to-point link if it is in full-duplex mode but is a shared link if it is in
half-duplex mode. RSTP fast convergence can only happen on point-to-point links and on edge ports.
This can expedite the convergence because this will have the port quickly transit to the forwarding state.
There are three option, auto, true and false, used to configure the type of the point-to-point link. If it is set
to auto, it means that RSTP will use the duplex mode resulting from auto-negotiation. This is the safest
method because although most links run in full-duplex mode, auto-negotiation may set it to half-duplex. If
it is set to true, the port will be treated as a point-to-point link and unconditionally transit to the forwarding
state. If it is set to false, fast transition to forwarding state will not happen on this port. Default: Auto
M Check:
Stands for Migration Check. Forces the port to send out an RSTP BPDU instead of a legacy
STP BPDU during the next transmission. The only benefit of this operation is to make the port quickly
revert to act as an RSTP port. Click the “M Check” button to send a RSTP BPDU from the specified port.
3.14 Trunking Configuration
The Port Trunking Configuration is used to configure the link aggregation settings on the switch. With link
aggregation, you can group multiple ports together that have the same speed, are full duplex and have
the same MAC address to be a single logical port, thus the combining the bandwidth of these ports. For
example, if there are three Fast Ethernet ports aggregated together into one logical port, then this logical
port has a bandwidth three times faster than a single Fast Ethernet port.
The switch supports two types of port trunking methods:
LACP:
Ports using Link Aggregation Control Protocol (according to the IEEE 802.3ad specification) as
their trunking method can choose their unique LACP Group ID (1-8) to form a logical “trunked port”. The
benefit of using LACP is that a port has to negotiate with its peer’s ports before it becomes a member of a
“trunk group” (also called aggregator). LACP is safer than the other trunking method, static trunks.
LACP does not support the followings:
•
Link Aggregation across switches.
•
Aggregation with non-IEEE 802.3 MAC links.
•
Operating in half-duplex mode.
•
Aggregating ports with different data rates.
Static Trunk:
Ports using Static Trunk as their trunk method can choose their unique Static Group ID
(also 1~8, this Static groupID can be the same with another LACP group ID) to form a logical “trunked
port”. The benefit of using Static Trunk is that a port can immediately become a member of a trunk group
without any handshaking with its peer. This is also a disadvantage because the peer ports connect to the
static trunk group may not know that they should be aggregated together to. Using Static Trunk on both
ends of a link is strongly recommended. Please also note that low speed links will be in a “not ready” state
when using static trunk to aggregate with high speed links.
The switch supports 8 “real trunked” groups. An LACP trunk group with more than one ready member-
ports is a “real trunked” group. An LACP trunk group with one or zero ready member-ports is not a “real
trunked” group. Any Static trunk group is a “real trunked” group. Each Trunking Group supports a
maximum of 12 ready member-ports. Please note that some decisions will automatically be made by the
system while you are configuring your trunking ports in order to conform to the rules of 802.3ad.
