79To do… Use the command… RemarksLocate the card problem for acard on a specified memberswitch (IRF mode)test diag-offline chassischassis-number slot slot-numberRequiredThe fault detection information will bedisplayed and saved in logs in theroot directory of the flash. Each log isnamed in the format ofdiag_slot_time, where slot indicatesthe location of the card and timeindicates when the fault detectionoperation is performed.CAUTION:• The AMB cannot be isolated.• If only one switching fabric module is working on the switch, it cannot be isolated.• Before upgrading a logic of the line card, keep the line card offline.• No other commands but the test diag-offline command can be executed for an isolated card.Otherwise, the configuration cannot take effect.NOTE:• Use the display device command to view whether a card is isolated, or, whether the card is in the offlinestate.• To minimize the interference on the system operation, force a switching fabric module that is operatingnormally offline before you unplug it.• When you execute the test diag-offline command, the path where the fault detection information issaved is also displayed following the fault detection information on the terminal. For example, you cansee flash:/diag_slot3_20080522_103458.txt.• Deliver the fault detection information to the H3C technical engineers or engineers of the sales agent.Configuring in-service hardware failure diagnosisand failure protectionA hardware failure may cause traffic forwarding failures and service interruption. To improve theautomatic failure detection and handling capabilities of the switch, you can configure in-servicehardware failure diagnosis and failure protection.The in-service hardware failure diagnosis and failure protection feature covers:• In-service hardware failure detection for chips, cards, and the forwarding service, and automatic fixactions taken for the detected failures.• Failure protection for ports: When a hardware failure is detected on a port, the switch automaticallyshuts down the port.• Failure protection for aggregation groups. When a hardware failure is detected on a member portof an aggregation group, if the member port is configured with the hardware-failure-protectionauto-down command, the port is shut down automatically; if the member port is not configured withthat command but it is not the last up port in the aggregation group, the port is also shut downautomatically; if the member port is not configured with that command and it is the last UP port inthe aggregation group, the port will not be shut down.
