Trapeze Networks MX-8-HA-LIC Scheda Tecnica
enabled on which access points throughout the
network. It is a holistic approach that is both
more efficient and more comprehensible.
network. It is a holistic approach that is both
more efficient and more comprehensible.
Once the service profiles have been defined and
created, they are pushed down to one primary
controller, known as the “seed” controller, which
in turn distributes the unified configuration to
all other controllers in the Virtual Controller
Cluster. Any controller may be the “seed”.
Primary and secondary “seed” controllers are
defined during the initial setup of the Virtual
Controller Cluster. Should the primary “seed”
fail, the secondary “seed” takes over.
created, they are pushed down to one primary
controller, known as the “seed” controller, which
in turn distributes the unified configuration to
all other controllers in the Virtual Controller
Cluster. Any controller may be the “seed”.
Primary and secondary “seed” controllers are
defined during the initial setup of the Virtual
Controller Cluster. Should the primary “seed”
fail, the secondary “seed” takes over.
Access Point Load Balancing
Members of the Virtual Controller Cluster col-
laborate with one another to intelligently deter-
mine the optimal balance for the operational
load. Working as a community, the controllers
in the cluster use load-balancing and capacity
planning algorithms to efficiently and dynami-
cally share access-point control.
laborate with one another to intelligently deter-
mine the optimal balance for the operational
load. Working as a community, the controllers
in the cluster use load-balancing and capacity
planning algorithms to efficiently and dynami-
cally share access-point control.
If a new controller is added to the cluster it is
equivalent to adding a slice of capacity to the
pool. Access Points simply get redistributed to
spread the load more evenly across the greater
pool of resources. Similarly, if a controller is
removed from the cluster - either for mainte-
nance or due to failure - the access points are
again redistributed for optimal loading.
equivalent to adding a slice of capacity to the
pool. Access Points simply get redistributed to
spread the load more evenly across the greater
pool of resources. Similarly, if a controller is
removed from the cluster - either for mainte-
nance or due to failure - the access points are
again redistributed for optimal loading.
Access Point Affinity
In the traditional hot-standby mode of redun-
dancy configuration, each access point “boots”
from, and is thereafter controlled by, a primary
controller, but also has a secondary controller
defined from which it will “boot” if the primary
controller is unavailable. This sense of having
“fixed” devices that an access point is tied to,
has its merits. For one thing, on a large campus
with more than one data center, or physically
distributed controllers, it lets administrators
keep traffic local and avoid wasting backbone
bandwidth with control-plane traffic. The High
Availability Module allows these benefits to be
preserved even in a clustering mode, by option-
ally allowing the APs to use a “preferred” group
of controllers from within the cluster.
dancy configuration, each access point “boots”
from, and is thereafter controlled by, a primary
controller, but also has a secondary controller
defined from which it will “boot” if the primary
controller is unavailable. This sense of having
“fixed” devices that an access point is tied to,
has its merits. For one thing, on a large campus
with more than one data center, or physically
distributed controllers, it lets administrators
keep traffic local and avoid wasting backbone
bandwidth with control-plane traffic. The High
Availability Module allows these benefits to be
preserved even in a clustering mode, by option-
ally allowing the APs to use a “preferred” group
of controllers from within the cluster.
NonStop Session Availability
In a traditional Hot-Standby redundancy con-
figuration, when a controller goes offline, so
do its APs. Even if APs are dual homed (either
directly or indirectly) to another controller,
they must drop all sessions, before reconnect-
ing to a “hot standby” controller. This typically
takes 15-20 seconds. Regardless how short the
boot time, for a user that was in a phone call,
or a doctor in the middle of reviewing patient
records, the interruption is more significant.
Many disrupted applications will require not
only re-authentication of the devices, but also
logging again at the application level.
figuration, when a controller goes offline, so
do its APs. Even if APs are dual homed (either
directly or indirectly) to another controller,
they must drop all sessions, before reconnect-
ing to a “hot standby” controller. This typically
takes 15-20 seconds. Regardless how short the
boot time, for a user that was in a phone call,
or a doctor in the middle of reviewing patient
records, the interruption is more significant.
Many disrupted applications will require not
only re-authentication of the devices, but also
logging again at the application level.
The Virtual Controller Cluster approach, avoids
these service interruptions altogether. In a con-
troller down scenario – any event that takes the
controller out of service - the many-to-many
redundancy allows the APs previously assigned
to the offline controller to be redistributed
among remaining controllers in the cluster,
instantaneously. Without losing a single active
session - not even voice calls.
these service interruptions altogether. In a con-
troller down scenario – any event that takes the
controller out of service - the many-to-many
redundancy allows the APs previously assigned
to the offline controller to be redistributed
among remaining controllers in the cluster,
instantaneously. Without losing a single active
session - not even voice calls.
Instant Adds Moves and Changes
Since the reliability is so robust, it means
administrators can perform in-service upgrades,
adds, moves and changes, at their convenience,
without any risk of downtime. There is no need
to schedule downtime - ever. This is a great
bonus for 24/7 operations such as hospitals and
factories.
administrators can perform in-service upgrades,
adds, moves and changes, at their convenience,
without any risk of downtime. There is no need
to schedule downtime - ever. This is a great
bonus for 24/7 operations such as hospitals and
factories.
When adds, moves and changes are effected,
there is no secondary maintenance to be per-
formed - not so with the hot-standy approach,
in which every change on a primary controller,
must also be replicated (and tested) on the
hot-standby.
there is no secondary maintenance to be per-
formed - not so with the hot-standy approach,
in which every change on a primary controller,
must also be replicated (and tested) on the
hot-standby.
2