Last updated: 27 August 2007
|
and TFT Color Display of the Myri-10G 10G-CLOS-ENCL Switch |
Home | Introduction | Web Interface | TFT Display | Appendix | Glossary
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.
This trap does not correspond to the 10G-POWER.
A non-zero value for this trap indicates that a crc error has been detected on a port of this crossbar.
This trap does not apply to the 10GXBar16.
This trap does not apply to the 10GXBar16.
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.
This trap does not apply to the 10GXBar16.
A non-zero value for this trap occurs when the synchronization for a xbar port fails. This trap corresponds to the local fault 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 port failed. This trap corresponds to the self test result counter for a specific sff port.
This state bit indicates that although the XAUI receiver has established a good link, the receiver on the other end of the link has not.
A non-zero value for this trap indicates that the hardware initialization test for this port failed. This trap corresponds to the self test result counter for a specific cx4 port.
A non-zero value for this trap indicates that the 10ge_port is down.
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.
This trap does not correspond to a 10GXBar16.
This trap does not apply to a 10GXBar16.
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 fault counter for a specific crossbar port.
This trap does not apply to the 10G-CLOS-ENCL.
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.
This state bit indicates that the XAUI receiver has not established a good link.
This counter indicates that a packet with an bad head byte has been received. It could be a byte without its high bit set, or a routing byte that routes to a nonexistent port (e.g. a routing byte of "-4" received on port 3 would route to port -1, which doesn't exist).
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 cx4 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.
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).
Internal Myricom use.
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 self test result counter reports if the self test of the cx4 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
port_state: 0 - ok
1 - testing/loading
2 - dead
This counter increments each time the port is rebooted.
Four bytes taken from the Lanai SRAM address 0 right before reboot.
led_info: bitwise status
0x04 - port up if set
0x02 - data sent
0x01 - data received
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.
This counter indicates that the XAUI receiver has not successfully synchronized the 4 XAUI lanes.
This counter indicates the number of times the receiver not aligned state bit has toggled between 0 and 1.
This counter indicates the total number of packets that have been received by this crossbar.
This counter indicates the number of packets that have been transmitted out of this crossbar.
This state bit indicates that a local synchronization failure has occurred.
This counter indicates the number of times the local fault state bit has toggled between 0 and 1.
This state bit indicates that a remote synchronization failure has occurred.
This counter indicates the number of times the remote fault 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 port down count counter indicates the total number of times that the value of the port down state bit has been toggled between 0 and 1.
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.
This state bit indicates if this power supply slot contains a power supply. 0 if not, and 1 if present.
This value indicates the maximum amount of power divided by 10 (in Watts) being drawn by the power supply. E.g., 85 for 850W.
This state bit indicates if the fan on the power supply has failed.
This counter indicates the numbers of times that the value of the fan bad state bit has toggled between 0 and 1.
Insufficient voltage was detected and the power supply was shutdown.
output undervoltage shutdown count
This counter indicates the number of times that the value of the output undervoltage shutdown state bit has toggled between 0 and 1.
The power supply shut down because it detected too high a voltage on its output.
output overvoltage shutdown count
This counter indicates the number of times that the value of the output overvoltage shutdown state bit has toggled between 0 and 1.
The power supply shut down because it detected an overload condition on its output.
output overload shutdown count
This counter indicates the number of times that the value of the output overload shutdown state bit has toggled between 0 and 1.
The power supply has overheated and shutdown.
overtemperature shutdown count
This counter indicates the number of times that the value of the overtemperature shutdown state bit has toggled between 0 and 1.
The power supply detected that its standby output (used only by the power supply monitoring circuitry) is not within specifications.
This counter indicates the number of times that the value of the standby voltage bad state bit has toggled between 0 and 1.
The power supply detected that its main output is not within specifications.
This counter indicates the number of times that the value of the output voltage bad 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 input bad state bit has toggled between 0 and 1.
The current speed (in RPM) of operation for this fan.
The minimum speed (in RPM) required for proper functioning of the fan.
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.
fan mode (1byte) - 0 if temperature is cool (slow fans)
1 if temperature is hot (fast fans)
Voltage on fan.
Last updated: 27 August 2007