Converging SAN and LAN Infrastructure with Fibre Channel over Ethernet for Efficient, Cost-Effective Date Centers

Converging SAN and LAN Infrastructure with Fibre Channel over Ethernet for Efcient, Cost-Effective Data Centers Page 5
Fibre Channel over Ethernet (FCoE) Facilitates consolidation of Local Area Network (LAN) and Storage Area Network
(SAN) trafc over a Unied Data Center Fabric
Data Center Bridging (DCB) Enables Ethernet fabric to support lossless trafc ows, making it suitable for SAN trafc
SAN connectivity to all servers FCoE enables all servers to access new or existing Fibre Channel SANs over standard
10GbE adapters, eliminating the need for Fibre Channel Host Bus Adapters and
secondary storage fabrics
Enhanced virtualization infrastructure Enables virtual machines to participate in live migration for load balancing and automatic
failover, as well as network-interface virtualization for enhanced trafc segmentation
Robust ecosystem of development
partnerships (including Cisco, Emulex,
Qlogic, Intel, and others)
Delivers integration of capabilities throughout the Unied Data Center Fabric hardware/
software solution stack
I/O consolidation and data center
network convergence
Reduces power and cooling requirements, for lower operating costs and
environmentally sound operation, in addition to lower infrastructure costs
Further, it provides a level of granularity that allows prioritization of
specic trafc types. The primary technologies that enable DCB are:
• Priority Flow Control allows denition of up to eight user
priorities on a single physical link, each of which has its own
logical lane that can be paused independently of the others.
This capability allows a port to send a Pause command with
less effect on the network as a whole, because it enables
more granular control over which trafc is affected, including
application of lossless transmission only to FCoE trafc.
• Congestion Notication responds to congestion by
communicating with upstream network devices, instructing
them to throttle specic transmissions to shape the trafc that
is causing congestion. That shaping is calculated to push
the congestion to the edge of the network and away from the
network core, limiting the effect of the congestion on network
performance overall.
• Enhanced Transmission Selection allocates bandwidth among
different trafc classes to help ensure appropriate priority and
latency for trafc that requires it (for example, to help ensure that
bandwidth requirements are met for storage trafc to help ensure
lossless transmission). The mechanism is exible enough
to allow other trafc classes to use idle allocated bandwidth,
helping ensure overall efcient use of network resources.
• Data Center Bridging Capability Exchange Protocol
(DCBCXP) is a management protocol that allows Enhanced
Ethernet to work transparently with conventional Ethernet by
dynamically discovering the capabilities of peer devices on
the network. For example, DCBCXP enables an edge switch
to discover the extent of the Enhanced Ethernet cloud and
the related capabilities of its peers, so that it knows how to
interact with them. DCBCXP also allows devices to verify
that conguration parameters such as user priorities are
compatible among devices and to push those parameters out
to peers as needed.
These capabilities enable a robust Enhanced Ethernet topology that
operates in parallel with traditional Ethernet over the same fabric for
maximum capability and exibility. Some of the capabilities provided
by a unied data center fabric are listed in Table 2.
Table 2. Unied data center fabric capabilities.