Last updated: 22 August 2007
This glossary contains a listing of all traps, counters, and state bits associated with the various switch components.
A non-zero value for this trap indicates the number of times that the microcontroller on this switch line card has been reset. This trap corresponds to the resets counter on the respective webpage for this switch line card.
A non-zero value for this trap indicates that an over-temperature shutdown has occurred for this switch line card. This trap corresponds to the overtemps counter on the switch line card. Examine the inlet and outlet temperatures, temperature 0 and temperature 1 respectively, for the indicated switch line card. The temperatures are reported in degrees Celsius. Shutdown occurs at 55 degrees Celsius, and normal operation can be restarted when the temperatures cools to 50 degrees Celsius or lower. Check for proper cooling and air flow around the switch, and also check for proper fan functioning. Once a switch line card had encountered an over-temperature shutdown, it will automatically power back on when the temperature drops to approximately 50 degrees Celsius.
A non-zero value for this trap can occur naturally when a switch line card is reseated, removed, or added. Also refer to the lost_communication trap.
A non-zero value for this trap indicates that the microcontroller on a specific switch line card has failed its self test. Other traps that are related to this situation are lost communication and reestablish_communication.
A non-zero value for this trap indicates that a fan tray has been removed from the switch chassis. This trap corresponds to the tray missing count counter for the fan tray in a specific slot of the switch.
A non-zero value for this trap indicates that a change has occurred in the ac power to the switch chassis. Perhaps there was a power fluctuation, or the electrical cord was accidentally unplugged, or a fuse was tripped. This trap corresponds to the ac bad count counter for a power supply.
A non-zero value for this trap indicates that a change has occurred in the dc current from the power supply in the switch to the switch components. If a power supply failed or was hot-swapped, then this trap would be set. This trap corresponds to the dc bad count counter for a power supply.
A non-zero value for this trap indicates that a crc error has been detected on a port of this crossbar.
A non-zero value for this trap indicates that a message has been routed to an out of range port on a crossbar or to a crossbar that is disconnected. This trap is related to routing, and corresponds to the illegal symbols counter for a specific crossbar port.
A non-zero value for this trap indicates that a message has been routed to an out of range port on a crossbar or to a crossbar that is disconnected. This trap is related to routing, and corresponds to the dead ports counter for a specific crossbar port. This trap may occur naturally only when the mapper is run.
A non-zero value for this trap occurs when a packet on its way into or out of a crossbar port is blocked, and the packet is dropped. This trap is related to blocking, and can be caused by deadlocks or insufficient receive buffers (software errors). This trap corresponds to the receive timeout count counter for a specific xbar port.
A non-zero value for this trap occurs when a cable is disconnected or there is some signal error on the cable. This trap corresponds to the overrun count counter for a specific crossbar port.
A non-zero value for this trap occurs when the synchronization for a xbar port fails. This trap corresponds to the local synch failure counter for a specific crossbar port.
A non-zero value for this trap occurs when a cable is disconnected or there is some signal error on the cable. It is normal for this trap to increase when a port is disconnected. This trap corresponds to the signal lost count counter for a specific quad port.
A non-zero value for this trap occurs when a cable is disconnected or there is some signal error on the cable. It is normal for this trap to increase when a port is disconnected. This trap corresponds to the signal lost count counter for a specific sff port.
A non-zero value for this trap indicates that the hardware initialization test for this crossbar failed. This trap corresponds to the self test result counter for a specific xbar port.
A non-zero value for this trap indicates that the hardware initialization test for this crossbar failed. This trap corresponds to the self test result counter for a specific sff port.
A non-zero value for this trap indicates that the microcontroller on the switch line card has encountered a software problem (watchdog timeout, illegal address, illegal opcode reset). One possible cause is an incorrect topology. This trap corresponds to the firmware faults counter for a specific switch line card.
A non-zero value for this trap can occur naturally when a switch line card is reseated, removed, or added. This trap corresponds to the lost communication count counter for a specific switch line card.
A non-zero value for this trap indicates that the switch line card in this slot of the switch chassis has changed. This trap corresponds to the changes counter for a specific slot in the switch chassis.
A non-zero value for this trap indicates that a fan has failed. This trap corresponds to the bad count counter for a specific fan.
A non-zero value for this trap indicates the value of fan_tray_missing_count if the fan tray is not currently missing, or fan tray missing = - 1 if the fan tray has not been reinstalled yet.
A non-zero value for this trap indicates that a change has occurred in the ac power to the switch chassis. Perhaps there was a power fluctuation, or the electrical cord was accidentally unplugged and replugged, or a fuse was tripped and reset.
A non-zero value for this trap indicates that a change has occurred in the dc current from the power supply in the switch to the switch components. If a failed power supply were hot-swapped, then this trap would be set.
A non-zero value for this trap occurs when a message has been routed to an out of range port on a xbar or to a xbar port that is disconnected. This trap may occur naturally only when the mapper is run. This trap corresponds to the invalid routes counter for a specific xbar port.
A non-zero value for this trap means that the port did not receive a heartbeat within 10 usecs. It is normal for the value of this trap to increase when a port is disconnected. This trap corresponds to the missed beats counter for a specific xbar port.
A non-zero value for this trap indicates the state of this port has changed. Perhaps it was previously shutdown and has now been reset.
A non-zero value for this trap occurs when a packet on its way into or out of a port is blocked, and the packet is dropped. This trap is related to blocking, and corresponds to the transmit timeout count counter for a specific xbar port.
A non-zero value for this trap indicates that the port has shut itself down due to illegal symbols, missed beats, or ... . This trap corresponds to port down counter for a specific xbar port.
A non-zero value for this trap occurs when a synchronization for a xbar port fails. This trap corresponds to the remote synch failure counter for a specific crossbar port.
A non-zero value for this trap indicates a failure of the quad fiber transceiver module. This trap corresponds to the module fault count counter for a specific quad port.
A non-zero value for this trap indicates a power fluctuation of the electrical current to a specific component on the switch.
overcurrent: 0 if circuit breaker hasn't tripped
1 if circuit breaker has tripped
A non-zero value for this trap indicates that the hardware initialization test for this crossbar failed. This trap corresponds to the self test result counter for a specific quad port.
The lost communication state bit is set to 1 when a switch line card is reseated or removed; i.e., the communication with the microcontroller on the switch line card has been lost.
This counter indicates the number of times that the value of the lost communication state bit has toggled between 0 and 1.
The changes counter indicates that the switch line card in this specific slot of the switch chassis has changed.
The state state bit indicates the operational state of the switch line card. Possible values are:
state: monitor -- The device is in the monitor state. This can
be due to an interrupted programming attempt,
a firmware crash, etc. When in this state,
the rest of the status information is NOT the
expected status.
selftest -- The device is doings its self-test. The rest
of the status information may be invalid.
overtemp -- The device is shutdown. The rest of the status
information may be invalid.
ok -- The device is operating. All status information
is valid.
dead -- A fatal failure has been detected, and the device
is shutdown. At least the self-test results
should be valid.
shutdown -- The board has been shutdown by writing a nonzero
value to the shutdown register. The rest of the
status information may be invalid.
master shutdown -- Only used on backplane microcontrollers,
indicating that they have been shutdown by the
master crossbar. The rest of the status
information may be invalid.
only master dead -- Only used on master backplane microcontrollers,
indicating that the crossbar controlled by the
master microcontroller is dead, but other
crossbars may be okay.
The resets counter indicates the number of times that the microcontroller on the switch line card has been reset or has had a software problem.
The firmware faults counter indicates that the microcontroller on the switch line card has encountered a software problem (watchdog timeout, illegal address, illegal opcode reset). One possible cause for a non-zero value of this counter is an incorrect topology. Check that the switch line cards are inserted in the proper slots.
For further details, refer to the specifications for the M3-CLOS-ENCL and M3-SPINE-ENCL switches, as well as this FAQ entry.
The overtemps counter indicates the number of times that this switch line card has overheated and shut itself down. A non-zero value of this counter indicates that the switch line card reached a temperature near 55 degrees Celsius (check the values of temperature 0 and temperature 1 on the Status - > Slot X webpage). This trap may be due to insufficient air conditioning or fan speed (check for fan faults in slot 26 and slot 27 in the web output).
This state bit is set to 1 when a cable is disconnected or there is some signal error on the cable.
This counter indicates the number of times that the value of the signal lost state bit has toggled between 0 and 1. It is normal for this counter to increase when a port is disconnected.
The module fault state bit refers only to a quad port and is set to 1 when a failure has occurred in the fiber transceiver module.
This counter indicates the number of times that the value of the module fault state bit has toggled between 0 and 1.
The self test result counter reports if the self test of the sff port, quad port, or xbar port was successful.
The possible values for this counter are as follows:
self test result: 0 if OK
1 if test is in progress
2 if failure
The good crcs indicates the number of packets that have been transmitted through this crossbar since the last time the crossbar was rebooted.
The bad crc trap occurs whenever a crc error has been detected on a port.
The illegal symbols trap is related to routing. This trap occurs when a message has been routed to an out of range port on a xbar or to a xbar port that is disconnected. illegal symbols are errors at the protocol level.
The missed beats trap means that the port did not receive a heartbeat within 10 usecs. It is normal for this counter to go up when a port is disconnected. Is the cable disconnected or some type of signal error on the cable?
The invalid routes trap is related to routing. This trap occurs when a message has been routed to an out of range port on a xbar or to a xbar port that is disconnected. This trap may occur naturally only when the mapper is run.
The dead ports counter is related to routing. It occurs when a message has been routed to an out of range port on a xbar or to a xbar port that is disconnected, and may occur naturally only when the mapper is run.
The data in counter indicates the total number of packets that have been received by this crossbar.
The data out counter indicates the number of packets that have been transmitted out of this crossbar.
The local synch failure state bit indicates that the synchronization of a xbar port has failed.
This counter indicates the number of times that the value of the local synch failure state bit has toggled between 0 and 1.
The remote synch failure state bit indicates that the synchronization of a xbar port has failed.
This counter indicates the number of times that the value of the remote synch failure state bit has toggled between 0 and 1.
The port down state bit indicates that a port has shut itself down due to illegal symbols, missed beats, or ...
The receive timeout state bit is related to blocking. When a packet on its way into or out of a port is blocked, this state bit is set to 1, and the packet is dropped. This is caused by deadlocks or insufficient receive buffers (software errors).
This counter indicates the number of times that the value of the receive timeout state bit has toggled between 0 and 1.
The transmit timeout state bit is related to blocking. When a packet on its way into or out of a port is blocked, this state bit is set to 1, and the packet is dropped. This is caused by deadlocks or insufficient receive buffers (software errors).
This counter indicates the number of times that the value of the transmit timeout state bit has toggled between 0 and 1.
The overrun state bit is set to 1 when a cable is disconnected or there is some signal error on the cable.
This counter indicates the number of times that the value of the overrun state bit has toggled between 0 and 1.
This state bit indicates that a change has occurred in the ac power to the switch chassis. It will be set to 1 if, for example, there was a power fluctuation, or the electrical cord was accidentally unplugged, or a fuse was tripped.
This counter indicates the number of times that the value of the ac bad state bit has toggled between 0 and 1.
This state bit indicates that a change has occurred in the dc current from the power supply in the switch to the switch components. If a power supply failed or was hot-swapped, then this state bit would be set to 1.
This counter indicates the number of times that the value of the dc bad state bit has toggled between 0 and 1.
The bad state bit is set to 1 if the fan has failed, and 0 if the fan is functioning properly.
This counter indicates the number of times that the value of the bad state bit has toggled between 0 and 1.
The tray missing state bit is set to 1 if a fan tray has been removed or not functional, and 0 if the fan tray is installed and at least one fan is working. All fans in a single fan tray will have the same value for this state bit.
This counter indicates the number of times that the value of the tray missing state bit has toggled between 0 and 1.
Last updated: 22 August 2007