Specifications

DATA CENTER BEST PRACTICES
SAN Design and Best Practices 8 of 84
Backup and disaster tolerance
•Is there a centralized backup? (This will determine the number of ISLs needed to minimize congestion at
peak loads.)
•What is the impact of backup on latency-sensitive applications?
•Is the disaster solution based on a metro Fibre Channel (FC) or Fibre Channel over IP (FCIP) solution?
Diagnostics and manageability
•What is the primary management interface to the SAN (Command-Line Interface [CLI], Brocade Network
Advisor, or third-party tool)?
•How often will Brocade FOS be updated?
•How will you validate cable and optics integrity?
Investment protection
•Support for future FC technology and interoperability
•Support for alternative technologies like FCoE (FC over Ethernet)
Operational Considerations
While Brocade fabrics are scalable in terms of port density and performance, the design goal should be to
ensure simplicity for easier management, future expansion, and serviceability (for example, use edge-core-edge
to a collapsed core; do not use both Inter-Fabric Routing IFR and Virtual Fabric (VF) if not really needed; turn on
port monitoring and fencing parameters for critical application).
Note: Brocade has tested sharing of 64 devices per Logical SAN (LSAN) zone and 12 Fibre Channel Routers
(FCRs) per backbone fabric with virtual fabrics. Any requirements beyond the tested conguration should be pre-
tested in a non-production environment, or you should actively monitor system resources like CPU and memory
utilization to minimize fabric anomalies.
Be the Pilot
Whether building a new SAN or connecting to an existing SAN, pre-staging and validating a fabric/application
prior to putting it into production ensures there are baseline metrics in terms of rated throughput, latency, and
expected Cyclic Redundancy Check (CRC) errors based on patch panel and physical cable infrastructure.
Diagnostics—Gen 5 Fibre Channel Platforms Only
For SANs deployed with Brocade Gen 5 Fibre Channel platforms, adapters, and optics, use the new ClearLink
diagnostics port type, or D_Port, to run diagnostics during pre-deployment or when there are susceptible physical
layer issues. Part of Brocade Fabric Vision technology, ClearLink is an ofine diagnostics tool that allows users
to automate a battery of tests to measure and validate latency and distance across the switch links. ClearLink
diagnostics can also be used to verify the integrity of all 16 Gbps transceivers in the fabric on a one-by-one
basis. In addition, a ClearLink diagnostic port requires only the individual ports that are attached to the link
being tested to go ofine, while leaving the rest of the ports to operate online, in isolation from the link. It can
also be used to test links to a new fabric switch without allowing the new switch to join or even be aware of the
current fabric. This fabric based physical layer validation enables the following:
•Local and long-distance measurements (5-meter granularity for 16 Gbps Small Form-factor Pluggables [SFPs]
and 50 meters for 10 Gbps SFPs)
•Latency measurements
•Link performance
•Transceiver health check
•Transceiver uptime