Specifications

DATA CENTER BEST PRACTICES
SAN Design and Best Practices 55 of 84
Reviewing performance requirements:
•Host-to-storage port fan-in/out ratios
•Oversubscription ratios:
– Host to ISL
– Edge switch to core switch
– Storage to ISL
•Size of trunks
•Routing policy and currently assigned routes; evaluate actual utilization for potential imbalances
Watching for latencies such as these:
•Poor storage performance
•Overloaded hosts or applications
•Distance issues, particularly changes in usage (such as adding mirroring or too much workload)
•Deal with latencies immediately; they can have a profound impact on the fabric.
In summary, although Brocade SANs are designed to allow for any-to-any connectivity, and they support provision-
anywhere implementations, these practices can have an adverse impact on the performance and availability
of the SAN if left unchecked. As detailed above, the network needs to be monitored for changes and routinely
evaluated for how well it meets desired redundancy and resiliency requirements.
SUPPORTABILITY
Supportability is a critical part of deploying a SAN. Follow the guidelines below to ensure that the data needed
to diagnose fabric behavior or problems have been collected. While not all of these items are necessary, they
are all pieces in the puzzle. You can never know which piece will be needed, so having all of the pieces available
is best.
•Congure Brocade Fabric Watch monitoring: Leverage Brocade Fabric Watch to implement proactive monitoring
of errors and warnings such as CRC errors, loss of synchronization, and high-bandwidth utilization.
•Congure syslog forwarding: By keeping historical log messages and having all switch messages sent to one
centralized syslog server, troubleshooting can be expedited and simplied. Forwarding switch error messages
to one centralized syslog server and keeping historical log messages enables faster and more effective
troubleshooting and provides simple monitoring functionality.
•Back up switch congurations: Back up switch congurations on a regular basis to so that you can restore
switch conguration in case a switch has to be swapped out or to provide change monitoring functionality.
•Follow Brocade best practices in the LAN infrastructure: Brocade best practices in the LAN infrastructure
recommend a setup of different physical LAN broadcast segments, for example, by placing IP routers between
segments or conguring different VLANs for the management interfaces of two fabric switches
•Enable audit functionality: To provide audit functionality for the SAN, keep track of which administrator made
which changes, usage of multiple user accounts (or RADIUS), and conguration of change tracking or audit
functionality (along with use of errorlog/syslog forwarding).
•Congure multiple user accounts (LDAP/OpenLDAP or RADIUS): Make mandatory use of personalized user
accounts part of the IT/SAN security policy, so that user actions can be tracked. Also, restrict access by
assigning specic user roles to individual users.
•Establish a test bed: Set up a test bed to test new applications, rmware upgrades, driver functionality, and
scripts to avoid missteps in a production environment. Validate functionality and stability with rigorous testing
in a test environment before deploying into the production environment.