38 Intel® Blade Server Ethernet Switch Module IXM5414E
Quality of Service (QoS)
The Quality of Service (QoS) features of the Intel® Blade Server Ethernet Switch Module
IXM5414E allow you to allocate network bandwidth according to the needs of the network users.
This section will give you an overview of the methods available.
Quality of Service technologies are intended to provide guaranteed, timely, delivery of specific
application data to a particular destination. In contrast, standard IP-based networks are designed to
provide “best effort” data delivery service. Best effort service implies that the network will attempt
to deliver the data in a timely fashion, although there is no guarantee. During times of congestion,
packets may be delayed, sent sporadically or dropped. For typical Internet applications, such as
electronic mail and file transfer, a slight degradation in service is acceptable and in many cases is
unnoticeable. Conversely, any degradation of service has undesirable effects on applications with
strict timing requirements, such as voice or multimedia.
QoS is a means of providing consistent, predictable data delivery by distinguishing packets that have
strict timing requirements from those that are more tolerant of delay. Packets with strict timing
requirements are given special treatment in a QoS-capable network. To accomplish this, all elements
of the network must be QoS-capable. If one node is unable to meet the necessary timing
requirements, this creates a deficiency in the network path and the performance of the entire packet
flow is compromised.
Bandwidth provisioning
Bandwidth provisioning allows you to deliver varying levels of allocated bandwidth to users sharing
the same physical interface. By mapping a subscriber’s traffic profile to a predefined policy and then
actively provisioning the maximum bandwidth consumed by that subscriber, you can provide
enhanced service offerings to your users. Bandwidth provisioning reduces the risk of network
congestion and prevents a small number of applications or users from consuming all the available
bandwidth.
Bandwidth provisioning provides Maximum Burst Rate (MBR) management for an interface and a
flexible framework for defining and extending traffic classes. It allows you to allocate bandwidth by
mapping a subscriber’s traffic profile (e.g. source/destination IP address, traffic type) to a prescribed
policy. Bandwidth provisioning actively provisions maximum bandwidth. For example, bandwidth
provisioning can enable monitoring and management of bandwidth for VLAN traffic based on
VLAN class IDs over an interface.
To run bandwidth provisioning you need to define Bandwidth Allocation Profiles (BAPs) and
Traffic Classes (TCs), and then associate the two:
Bandwidth Allocation Profile
A transmission link definition which specifies a Bandwidth Bucket Identifier, as
well as maximum bandwidth allowances.
Traffic Class The definition of the traffic to which a set of rules will apply. A class is defined by
specifying a VLAN Identifier and an interface number, along with the class priority.
A default BAP, which you cannot modify, is assigned to all new TCs. Any given BAP may be
assigned to multiple TCs. Once you have defined the BAPs and TCs, and attached BAPs to the TCs,
VLAN traffic on the specified interfaces will not exceed the maximum configured bandwidth.