HP StorageWorks 2/16 - SAN Switch Manual page 244

Fabric Watch User Guide
You can also use the portshow command to view the port health that Fabric Watch
maintains. Valid port states are Healthy, Marginal, Faulty, and Offline.
Port Error Persistence
Port states indicate the overall health of a switch. While a port is operating, Fabric Watch
monitors run every six seconds, gathering measurements about the port and looking for any
abnormal behavior. When a value goes out of range, the monitor triggers.
Whenever a monitored value deviates from a user-defined range, the state of the port changes
from a healthy state to an abnormal state. A port must remain in the abnormal state for a
user-defined period of time before the port state affects the overall switch state. The monitor
returns to a normal state when a problem is temporary. However, if the problem persists for 18
seconds or more, the state of the port changes from Healthy to Marginal. The port state
remains Marginal as long as the problem exists.
You can modify the default error persistence value of 18. To configure the port error
persistence time, issue the fwset --port --persistence command.
244 Fabric OS 3.x Document Addendum