Specifications

DATA CENTER BEST PRACTICES
SAN Design and Best Practices 27 of 84
IOPS and VMs
Another method for determining bandwidth and/or oversubscription is to use the IOPS between host and
storage devices. If the typical I/O size is known, along with the typical number of IOPS, then the administrator
can calculate both average and estimated peak loads in terms of Megabytes per second (MB/sec). Next, look
at the paths through the network for these I/Os, along with I/Os from other devices using the same network
paths. Then use these data points to calculate bandwidth utilization and/or oversubscription ratios for devices
and ISLs.
The use of VMs and the mobility of these VMs can make such IOPS calculations a challenge, as loads can shift
when VMs move. Thus, the administrator needs to be aware of the potential VM loads on each physical server
and their associated application loads for VMs.
While these calculations can certainly produce an accurate picture of bandwidth requirements for the storage
network, they can be complicated even in a small network topology. This is why the simple approach discussed
above is generally recommended.
ROUTED TOPOLOGIES—METASANS
The FC-FC routing service enables Fibre Channel SANs to share devices between two or more fabrics without
merging those fabrics. The advantages for a routed topology are a reduced number of switch domains and
zones for management within an edge fabric, fault isolation to a smaller edge fabric, interoperability with legacy
fabrics, and increased security. In general, edge fabrics with Fibre Channel Routing (FCR) topologies follow the
same basic best practice design guidelines as traditional fabrics, core-edge architectures for example. The FCR
feature can be used for local or between fabrics across dark ber or Wide-Area Networks (WANs) using FCIP.
Note: Refer to Brocade Scalability Guidelines for FCR scalability limits.
The primary considerations for using FCR are as follows:
•A limited number of LUNs shared between fabrics
•A limited number of servers that need to share LUNs between fabrics
•Share archiving devices like tape libraries
•The migration of legacy M-EOS or Brocade FOS fabrics to current Brocade FOS-based platforms
•OEM support
•Security separation in a managed services environment
There should be redundancy at the fabric, switch, and Inter-Fabric Link (IFL) levels (see Figures 17–19). A routed
SAN, or MetaSAN, environment consists of multiple edge fabrics interconnected through one or more backbone
fabrics. Multiple backbone fabrics are in parallel and belong to only the A or B fabric, not both. A core-edge
topology can be used at both the backbone level and at the edge fabric level, such that edge fabrics and
backbone fabrics are both deployed in a core-edge fashion.