Using Error Log and Trace Facilities in Sequence

Fibre Channel Adapter/1063 User's Guide and Reference

Using Error Log and Trace Facilities in Sequence

The following sequence gathers the majority of information required by support personnel to begin an investigation of a problem:

Generate the error log report by entering the following command on the command line:
```
errpt -a > /tmp/errlog
```
Examine the error log report for FC entries and follow the recommended actions for those entries. See "FC Error Log Detail Data" for more information.
Start the trace facility with the following command:
```
trace -a -j "310 311 312 313 314 315 316 317 318"
```
If possible, re-create the problem.
Stop the trace facility with the following command:
```
trcstop
```
Generate the trace report with the following command:
```
trcrpt > /tmp/trace.rpt
```
Examine the data in the report.
Examine the error log report for additional FC entries generated during the problem re-creation and follow the recommended actions for those entries.
If the problem continues, report it to the appropriate service personnel. Include the following in your report:
- Detailed information about how to re-create the problem
- Error log report (output of the errpt command)
- Data from the trace report (output of the trcrpt command)
- Actions taken and their results.

FC Error Log Detail Data

All error log entries contain a copy of the fcs_error_log structure from the /usr/include/sys/fcs_user.h include file in the detail data section of every error log. The following detail data fields are defined:

`state`	Indicates the state of this device instance.
`adap_cmd`	Indicates the number of commands pending for the adapter.
`open_status`	Indicates the status of the open request.
`num_cancels`	Indicates the number of outstanding cancel elements.
`num_pools`	Indicates the number of outstanding pool elements.
`nport_addr`	Indicates the nport_address of the card.
`cbsp_value`	Indicates the last CBSP value.
`isp_value`	Indicates the last ISP value.
`mc_err`	Indicates the last Micro Channel adapter bus error.
`pio_err`	Indicates the last programmed input/output (PIO) exception.
`limbo_reason`	Indicates the last limbo entry condition.
`bugout_reason`	Indicates the last bugout entry condition.
`adap_data[64]`	Indicates adapter-specific log data.

FC Errors

The FC device driver logs the following types of errors :

PIO operational errors
Device hardware failures
Direct memory access (DMA) operational errors
Network errors
Other device-specific errors

Possible error values include:

+FC_960_FAULT	Indicates there was a failure with the I960 processor on the adapter.
+FC_PERM_DMA_ERROR	Indicates there was an unrecoverable DMA error.
+FC_TEMP_DMA_ERROR	Indicates there was a DMA error, but the driver was able to recover from the error.
+FC_UNKNOWN_ERROR	Indicates there was an unknown hardware error on the adapter.
+FC_LINK_DOWN	Indicates the fibre-optic link is disconnected or not functioning.
+FC_LINK_UP	Indicates the fibre-optic link has been re-established.
+FC_LIMBO_EXIT	Indicates the device driver has successfully exited the error recovery procedure.
+FC_PIO_FAIL	Indicates the FC adapter driver had a PIO failure and is retrying the operation.
+FC_PIO_DEAD	Indicates the FC driver has retried a failed PIO operation three times and is shutting down the device.
+FC_CMD_TIMEOUT	Indicates the FC adapter driver has timed out on a command issued to the adapter. The FC adapter driver will log error recovery on this condition.

+FC_960_FAULT

The following fields are defined in the error record template for 960-fault errors:

`Class`	Equals a value of `H` , which indicates a hardware error.
`Err_Type`	Equals a value of `Perm` , which indicates a permanent failure.
`Report`	Equals a value of `True` , which indicates this error should be included when an error report is generated.
`Log`	Equals a value of `True` , which indicates an error log entry should be created when this error occurs.
`Alert`	Equals a value of `False` , which indicates this error should not be forwarded to the Network Alert Manager.
`Err_Desc`	Equals a value of `1010` , which indicates an adapter error.
`Prob_Causes`	Equals a value of `3330` , which indicates the adapter hardware.
`Fail_Causes`	Equals a value of `3300` , which indicates an adapter failure.
`Fail_Actions`	Equals the following values: `E800` , which indicates system diagnostics should be run. `F011` , which indicates no further action is required unless the problem persists.

+FC_PERM_DMA_ERROR

The following fields are defined in the error record template for permanent DMA errors:

`Class`	Equals a value of `H` , which indicates a hardware error.
`Err_Type`	Equals a value of `Perm` , which indicates a permanent failure.
`Report`	Equals a value of `True` , which indicates this error should be included when an error report is generated.
`Log`	Equals a value of `True` , which indicates an error log entry should be created when this error occurs.
`Alert`	Equals a value of `False` , which indicates this error should not be forwarded to the Network Alert Manager.
`Err_Desc`	Equals a value of `EB75` , which indicates a DMA error.
`Prob_Causes`	Equals a value of `EB0A` , which indicates the Micro Channel bus.
`Fail_Causes`	Equals a value of `EA5A` , which indicates a Micro Channel bus failure.
`Fail_Actions`	Equals a value of `E800` , which indicates system diagnostics should be run.

+FC_TEMP_DMA_ERROR

The following fields are defined in the error record template for temporary direct memory access (DMA) errors:

`Class`	Equals a value of `H` , which indicates a hardware error.
`Err_Type`	Equals a value of `Temp` , which indicates a temporary failure.
`Report`	Equals a value of `True` , which indicates this error should be included when an error report is generated.
`Log`	Equals a value of `True` , which indicates an error log entry should be created when this error occurs.
`Alert`	Equals a value of `False` , which indicates this error should not be forwarded to the Network Alert Manager.
Indicates a temporary, adapter-detected DMA error.
`Err_Desc`	Equals a value of `EB75` , which indicates a DMA error.
`Prob_Causes`	Equals a value of `EB0A` , which indicates the Micro Channel bus.
`Fail_Causes`	Equals a value of `EA5A` , which indicates a Micro Channel bus failure.
`Fail_Actions`	Equals a value of `F011` , which indicates no further action is required unless the problem persists.

+FC_UNKNOWN_ERROR

The following fields are defined in the error record template for unknown errors:

`Class`	Equals a value of `H` , which indicates a hardware error.
`Err_Type`	Equals a value of `Perm` , which indicates a permanent failure.
`Report`	Equals a value of `True` , which indicates this error should be included when an error report is generated.
`Log`	Equals a value of `True` , which indicates an error log entry should be created when this error occurs.
`Alert`	Equals a value of `False` , which indicates this error should not be forwarded to the Network Alert Manager.
`Err_Desc`	Equals a value of `1010` , which indicates an adapter error.
`Prob_Causes`	Equals the following values: `3330` , which indicates the adapter hardware. `3331` , which indicates the adapter microcode.
`Fail_Causes`	Equals a value of `3300` , which indicates an adapter failure.
`Fail_Actions`	Equals the following values: `E800` , which indicates system diagnostics should be run. `F011` , which indicates no further action is required unless the problem persists.

+FC_LINK_DOWN

The following fields are defined in the error record template for link-down errors:

`Class`	Equals a value of `H` , which indicates a hardware error.
`Err_Type`	Equals a value of `Temp` , which indicates a temporary failure.
`Report`	Equals a value of `True` , which indicates this error should be included when an error report is generated.
`Log`	Equals a value of `True` , which indicates an error log entry should be created when this error occurs.
`Alert`	Equals a value of `False` , which indicates this error should not be forwarded to the Network Alert Manager.
`Err_Desc`	Equals a value of `EB76` , which indicates a fibre-channel link is down.
`Prob_Causes`	Equals the following values: `3330` , which indicates the adapter hardware. `3400` , which indicates a cable. `EB0B` , which indicates the fibre-channel switch.
`Fail_Causes`	Equals the following values: `EA03` , which indicates a fibre-optic cable failure. `3330` , which indicates an adapter failure. `3400` , which indicates failure of a loose or defective cable. `EA5B` , which indicates a fibre-channel switch failure.
`Fail_Actions`	Equals a value of `0302` , which indicates that cables and their connections should be checked.

+FC_LINK_UP

The following fields are defined in the error record template for link-up errors:

`Class`	Equals a value of `H` , which indicates a hardware error.
`Err_Type`	Equals a value of `Temp` , which indicates a temporary failure.
`Report`	Equals a value of `True` , which indicates this error should be included when an error report is generated.
`Log`	Equals a value of `True` , which indicates an error log entry should be created when this error occurs.
`Alert`	Equals a value of `False` , which indicates this error should not be forwarded to the Network Alert Manager.
`Err_Desc`	Equals a value of `EB77` , which indicates a fibre-channel link is up.
`Prob_Causes`	Equals the following values: `3330` , which indicates the adapter hardware. `3400` , which indicates a cable. `EA5B` , which indicates the fibre-channel switch.
`Fail_Causes`	Equals the following values: `EA03` , which indicates a fibre-optic cable failure. `3300` , which indicates an adapter failure. `3400` , which indicates failure of a loose or defective cable. `EA5B` , which indicates a fibre-channel switch failure.
`Fail_Actions`	Equals a value of `F011` , which indicates that no action is required unless the problem persists.

+FC_LIMBO_EXIT

The following fields are defined in the error record template for limbo-exit errors:

`Class`	Equals a value of `H` , which indicates a hardware error.
`Err_Type`	Equals a value of `Perm` , which indicates a permanent failure.
`Report`	Equals a value of `True` , which indicates this error should be included when an error report is generated.
`Log`	Equals a value of `True` , which indicates an error log entry should be created when this error occurs.
`Alert`	Equals a value of `False` , which indicates this error should not be forwarded to the Network Alert Manager.
`Err_Desc`	Equals a value of `6100` , which indicates a microcode program error.
`Prob_Causes`	Equals the following values: `3330` , which indicates the adapter hardware. `3331` , which indicates the adapter microcode.
`Fail_Causes`	Equals a value of `3300` , which indicates an adapter failure.
`Fail_Actions`	Equals a value of `F011` , which indicates no action is required unless the problem persists.

+FC_PIO_FAIL

The following fields are defined in the error record template for PIO failures:

`Class`	Equals a value of `H` , which indicates a hardware error.
`Err_Type`	Equals a value of `Perm` , which indicates a permanent failure.
`Report`	Equals a value of `True` , which indicates this error should be included when an error report is generated.
`Log`	Equals a value of `True` , which indicates an error log entry should be created when this error occurs.
`Alert`	Equals a value of `False` , which indicates this error should not be forwarded to the Network Alert Manager.
`Err_Desc`	Equals a value of `E874` , which indicates a PIO exception.
`Prob_Causes`	Equals the following values: `3330` , which indicates the adapter hardware. `0005` , which indicates the system I/O bus.
`Fail_Causes`	Equals a value of `0005` , which indicates a system I/O bus failure.
`Fail_Actions`	Equals a value of `E800` , which indicates system diagnostics should be run.

+FC_PIO_DEAD

The following fields are defined in the error record template for PIO-dead failures:

`Class`	Equals a value of `H` , which indicates a hardware error.
`Err_Type`	Equals a value of `Perm` , which indicates a permanent failure.
`Report`	Equals a value of `True` , which indicates this error should be included when an error report is generated.
`Log`	Equals a value of `True` , which indicates an error log entry should be created when this error occurs.
`Alert`	Equals a value of `False` , which indicates this error should not be forwarded to the Network Alert Manager.
`Err_Desc`	Equals a value of `E874` , which indicates a PIO exception.
`Prob_Causes`	Equals the following values: `3330` , which indicates the adapter hardware. `0005` , which indicates the system I/O bus.
`Fail_Causes`	Equals a value of `0005` , which indicates a system I/O bus failure.
`Fail_Actions`	Equals a value of `E800` , which indicates system diagnostics should be run.

+FC_CMD_TIMEOUT

The following fields are defined in the error record template for time-out errors:

`Class`	Equals a value of `H` , which indicates a hardware error.
`Err_Type`	Equals a value of `Perm` , which indicates a permanent failure.
`Report`	Equals a value of `True` , which indicates this error should be included when an error report is generated.
`Log`	Equals a value of `True` , which indicates an error log entry should be created when this error occurs.
`Alert`	Equals a value of `False` , which indicates this error should not be forwarded to the Network Alert Manager.
`Err_Desc`	Equals a value of `6100` , which indicates a microcode program error.
`Prob_Causes`	Equals the following values: `3330` , which indicates the adapter hardware. `3331` , which indicates the adapter microcode.
`Fail_Causes`	Equals a value of `3300` , which indicates an adapter failure.
`Fail_Actions`	Equals the following values: `E800` , which indicates system diagnostics should be run. `F011` , which indicates no further action is required unless the problem persists.

[ Previous | Contents | Home | Search ]