Users Guide

192 | Secure Enterprise Mesh Dell PowerConnect ArubaOS 5.0 | [User Guide
z Using a new mesh link if the current mesh link goes down
If an uplink goes down, the affected mesh nodes re-establish a connection with the mesh portal by re-scanning
to choose a new path to the mesh portal. If a mesh portal goes down, and a redundant mesh portal is available,
the affected mesh nodes update their forwarding tables to reflect the path to the new mesh portal.
Link Metrics
Mesh points use the configured algorithm to compute a metric value, or “path cost,” for each potential uplink
and select the one with the lowest value as the optimal path to the mesh portal.Table 38 describes the
components that make up the metric value: node cost, hop count, link cost and 802.11 capacity.
The link metrics indicate the relative cost of a path to the mesh portal. The best path (lowest metric value) is
used to create the uplink. The mesh portal advertises a cost of 0, while all other mesh nodes advertise a
cumulative cost based on the parent mesh node.
Optimizing Links
You can configure and optimize operation of the link metric algorithm via the mesh radio profile. These
configurable mesh link trigger thresholds can determine when the uplink or mesh path is dropped and another is
chosen, provide enhanced network reliability, and contain flapping links. Although you can modify the behavior
of the link metric algorithm, Dell recommends the default values for most deployments. For information, see
“Metric algorithm” on page200.
Mesh Profiles
Mesh profiles help define and bring-up the mesh network. The following sections describe the mesh cluster, mesh
radio, and mesh recovery profiles in more detail.
The complete mesh profile consists of a mesh radio profile, RF management (802.11a and 802.11g) radio profiles,
a high-througput SSID profile (if your deployment includes 802.11n-capable APs), a mesh cluster profile, and a
read-only recovery profile. The recovery profile is dynamically generated by the master controller; you do not
explicitly configure the recovery profile.
Dell provides a “default” version of the mesh radio, RF management, high-througput SSID and cluster profiles
with default values for most parameters. You can use the “default” version of a profile or create a new instance of
a profile which you can then edit as you need. You can change the values of any parameter in a profile. You have
Table 38 Mesh Link Metric Computation
Metric Description
Node cost Indicates the amount of traffic expected to traverse the mesh node. The more traffic, the higher the node cost.
When establishing a mesh link, nodes with less traffic take precedence. The node cost is dependent on the
number of children a mesh node supports. It can change as the mesh network topology changes, for example
if new children are added to the network or old children disconnect from the network.
Hop count Indicates the number of hops it takes the mesh node to get to the mesh portal.
The mesh portal advertises a hop count of 0, while all other mesh nodes advertise a cumulative count based
on the parent mesh node.
Link quality Represents the quality of the link to an active neighbor. The higher the Received Signal Strength Indication
(RSSI), the better the path to the neighbor and the mesh portal. If the RSSI value is below the configured
threshold, the link cost is penalized to filter marginal links. A less direct, higher quality link may be preferred
over the marginal link.
802.11 capacity High-throughput APs can send 802.11 information elements (IEs) in their management frames, allowing high-
througput mesh nodes to identify other mesh nodes with a high-througput capacity. High-throughput mesh
points prefer to select other 802.11-capable mesh points in their path to the mesh portal, but will use a legacy
path if no high-throughput path is available.