Brocade Fabric OS Administrator's Guide Supporting Fabric OS v6.3.0 (53-1001336-02, November 2009)

436 Fabric OS Administrator’s Guide
53-1001336-02
Bottleneck detection
18
Bottleneck detection
A bottleneck is a port in the fabric where frames cannot get through as fast as they should. In other
words, a bottleneck is a port where the offered load is greater than the achieved egress
throughput. Bottleneck detection does not require a license.
Bottlenecks can cause undesirable degradation in throughput on various links. When a bottleneck
occurs at one place, other points in the fabric can experience bottlenecks as the traffic backs up.
A latency bottleneck is a port where the offered load exceeds the rate at which the other end of the
link can continuously accept traffic, but does not exceed the physical capacity of the link. This
condition can be caused by a device attached to the fabric that is slow to process received frames
and send back credit returns. A latency bottleneck due to such a device can spread through the
fabric and can slow down unrelated flows that share links with the slow flow.
The bottleneck detection feature identifies devices attached to the fabric that are slowing down
traffic. The stress on the devices might be due to a large number of real or virtual machines
creating a large workload for the device. You can then investigate and optimize the resource
allocation for the device. Using the zone setup and Top Talkers, you can also determine which flows
are destined to the affected F_Ports.
This feature detects latency bottlenecks on F_Ports or FL_Ports and reports the bottlenecks
through RASlog alerts. You can set alert thresholds for the severity and duration of the bottleneck.
The bottleneck detection feature enables you to do the following:
• Prevent degradation of throughput in the fabric.
The bottleneck detection feature alerts you to the existence and locations of devices that are
causing latency. If you receive RASlog alerts for one or more F_Ports, use the CLI to check
whether these F_Ports have a history of bottlenecks.
• Reduce the time it takes to troubleshoot network problems.
If you notice one or more applications slowing down, you can determine whether any latency
devices are attached to the fabric and where. You can use the CLI to display a history of
bottleneck conditions on a port. If the CLI shows above-threshold bottleneck severity, you can
narrow the problem down to device latency rather than problems in the fabric.
NOTE
Best practice is to turn bottleneck detection on for targets, and leave it on.
Supported configurations for bottleneck detection
Note the following configuration rules for bottleneck detection:
• Bottleneck detection is supported only on Fibre Channel ports.
• Bottleneck detection is supported only on the following port types:
- F_Ports
- FL_Ports
• F_Port trunks are not supported.
• Bottleneck detection is supported in Access Gateway mode.