ALR0652E The Class 3 Send Timeout Frame Rate of port port name in switch switch name was measured to be measured value cnt/s. This violated the defined critical-stress boundary value of boundary value cnt/s.

Explanation

Class 3 Send Timeout Frame Rate measures the rate at which class 3 frames had to be discarded at the transmission port due to timeout caused by lack of buffer credits. Buffer to Buffer Credit is used as a mechanism to enforce flow-control by the fibre channel protocol, and ensures that a transmitting port which is sending data does not overrun the ability of the receiving port to receive that data. When out of buffer credits, the port waits for a configurable amount of time (e.g. 500 ms) to receive credit and send the waiting frame. If no credit is received, a timeout condition occurs and the waiting frame is discarded. Note that if egress and ingress ports are on the same blade, the Class 3 Receive Timeout Frame rate increases for the ingress port(s) at the same time the Class 3 Send Timeout Frame rate increases for the egress port.

A threshold was defined on this metric, which causes the measured value to be compared to the defined boundaries for each set of performance statistics collected from the switch. This message indicates that a threshold boundary violation occurred. In this case, the specified port in the switch has a Class 3 Send Timeout Frame Rate value that is greater than or equal to the critical-stress boundary that was defined for the related threshold.

Action

The threshold violation might indicate that a hardware or configuration problem exists either for this port or for the connected fabric. Isolated errors can be ignored in most cases, but if the rate remains consistently high over time, this may be a cause for concern. Follow the trouble-shooting guidelines for the switch, and for the connected entity (storage system, switch, or server HBA).

Because isolated temporary errors can be ignored in most cases, consider enabling the alert suppression option "Suppress alerts unless the triggering condition has been violated continuously for a specified length of time" for this threshold. This can greatly reduce the number of unnecessary alerts.