Specifications
DATA CENTER BEST PRACTICES
SAN Design and Best Practices 23 of 84
Fabric Watch tracks a variety of SAN fabric elements and events. Monitoring fabric-wide events, ports, and
environmental parameters enables early fault detection and isolation as well as performance measurement. You
can congure fabric elements and alert thresholds on an individual port basis, and you can also easily integrate
Fabric Watch with enterprise system management solutions.
Fabric Watch provides customizable monitoring thresholds. You can congure Fabric Watch to provide notication
before problems arise, such as reporting when network trafc through a port is approaching the bandwidth limit.
This information enables you to perform pre-emptive network maintenance, such as trunking or zoning, and
avoid potential network failures.
Fabric Watch lets you dene how often to measure each switch and fabric element and specify notication
thresholds. Whenever fabric elements exceed these thresholds, Fabric Watch automatically provides notication
using several methods, including e-mail messages, SNMP traps, and log entries.
Fabric Watch was signicantly upgraded starting in Brocade FOS v6.4, and it continues to be a major source
of early warning for fabric issues. Useful enhancements, such as port fencing to protect the fabric against
misbehaving devices, are added with each new release of Brocade FOS.
Brocade Fabric Watch Recommendations
Brocade Fabric Watch is an optional feature that provides monitoring of various switch elements. Brocade Fabric
Watch monitors ports based on the port type, for example, F_Port and E_Port classes, without distinguishing
between initiators and targets. Since the monitoring thresholds and desired actions are generally different for
initiators and targets, it is recommended that these devices be placed on different switches so that Brocade
Fabric Watch settings can be applied accordingly.
Note: For additional details, see the Brocade Fabric Watch Administrator’s Guide.
RAS Log
RAS log is the Brocade FOS error message log. Messages are organized by Brocade FOS component, and each
one has a unique identier as well as severity, source and platform information and a text message.
RAS log is available from each switch and director via the “errdump” command. RAS log messages can be
forwarded to a syslog server for centralized collection or viewed within Brocade Network Advisor via the
Master Log.
Audit Log
The Audit log is a collection of information created when specic events are identied on a Brocade platform.
The log can be dumped via the auditdump command, and audit data can also be forwarded to a syslog server
for centralized collection.
Information is collected on many different events associated with zoning, security, trunking, FCIP, FICON, and
others. Each release of the Brocade FOS provides more audit information.
Brocade SAN Health
Brocade SAN Health provides snapshots of fabrics showing information such as switch and rmware levels,
connected device information, snapshots of performance information, zone analysis, and ISL fan-in ratios.
Design Guidelines
Brocade strongly recommends implementing some form of monitoring of each switch. Often issues start out
relatively benignly and gradually degrade into more serious problems. Monitoring the logs for serious and error
severity messages will go a long way in avoiding many problems.
•Plan for a centralized collection of RAS log, and perhaps Audit log, via syslog. You can optionally lter these
messages relatively easily through some simple Perl programs.