Command ERRPT for viewing error logs on AIX

The following is a brief introduction to the error log reporting mechanism of AIX and the meanings of some parameters: i. Mechanism of error logs first we need to understand how error logs are saved and recognized by users. In fact, when a function module of the system detects an error event, it will be recorded in the/dev/error device, then, the error log process errdemon reads the error log from the/dev/error file and writes it to/var/adm/ras/errlog. For specific settings related to the error log, you can see through the/usr/lib/errdemon-l command: error Log Attributes upload Log File/var/adm/ras/errlogLog Size 106496 bytesMemory Buffer Size 32768 bytesDuplicate Removal trueDuplicate Interval 10000 millisecondsDuplicate Error Maximum 1000 <aix53testos1>: /# The Error Log File shown above is/var/adm/ras/errlog, which is also the default location for storing error logs. 2. View error logs: The errpt command errpt is the most familiar command for the AIX system administrator. The following is the output of an errpt command: <aix61haos1>: /# errptIDENTIFIER timestamp t c RESOURCE_NAME limit 0226155209 p s src software program ERROR12081DC6 0226155109 p s harmad software program ERROR12081DC6 0226155109 p s harmad software program limit 0224164009 I O SYSJ2 UNABLE TO ALLOCATE SPACE IN FILE limit 0220140909 I o grpsvcs Group Services daemon started97419D60 0220140909 I O topsv Cs Topology Services daemon started6D19271E 0220133009 I O topsvcs Topology Services daemon stopped the output here is divided into six columns: 1. error IDENTIFIER: it is not unique. It determines the error template used. Obviously, the same error IDENTIFIER is the same. 2. TIMESTAMP: the time when an error occurs. MMDDhhmmYY indicates the time of the month, day, and year in sequence. 3. TYPE: the TYPE of the error, or the severity of the error, is divided into six types: PEND device or function component may be missing short PPERF performance seriously decline PPERM hardware device or software module damage, the temporary PTEMP error has been diagnosed. after retrying, the normal TINFO message has been restored. If it is not an error, IUNKN cannot determine the severity of the error U 4. category class c: indicates the error source H hardware fault HardwareS software fault SoftwareO manual operation OperationU cannot be determined Unknown 5. resource Name RESOURCE_NAME the software or hardware with the wrong Resource Name is initially detected. This does not mean that the resource is faulty, but is only first discovered in it. 6. to describe the specific meaning of the error representative, call the IBM Support Hotline for help. 7. The errpt command generates an error report from the record in the error log, which contains the flag of the selected error that meets the specific criteria. By default, you can display the records of error logs in the reverse order of errors and records. With the-c (parallel) flag, you can display these errors when errors occur. If the-I flag is not used together with the errpt command, the error log file processed by errpt is the one specified in the error log configuration database. (To view the information in the error log database, use the errrdemon command .) The default Summary Report contains a row of data for each error. You can use a flag to generate reports of different formats. Note: The errpt command does not analyze error logs. To analyze the error logs, use the diag command. However, when the error log analysis is executed, the diagnosis program will add the diagnosis information back to the error log, which will appear after the detailed information data of the corresponding error log record. Example: 1. to display a complete summary report, enter errpt 2. to display a complete detailed report, enter errpt-a 3. to display all error reports recorded by the error code E19E094F, enter errpt-a-j E19E094F 4. to display a detailed report of all errors recorded in the past 24 hours, enter errpt-a-s mmddhhmmyy. Here, the mmddhhmmyy string is equal to the current month, day, and hour, minute and year minus 24 hours. 5. to list the error log templates whose logs are disabled by any error log, enter errpt-t-F log = 0 6. if you want to import data from/var/adm/ras/errlog. to view all records in the alternate error log file, enter errpt-I/var/adm/ras/errlog. alternate 7. if you want to import data from/var/adm/ras/errlog. to view all hardware records in the alternate error log file, enter errpt-I/var/adm/ras/errlog. alternate-d H 8. to display detailed reports of all errors recorded by the error label ERRLOG_ON, enter errpt-a-J ERRLOG_ON 9. to display detailed reports of all errors and group repeated errors, enter errpt-aD. 3. Clear error logs by running the errclear command. In fact, manual cleanup is not required. After the system is installed, the script that is automatically executed regularly contains the following lines: 0 11 ***/usr/bin/errclear-d S, O 300 12 ***/usr/bin/errclear-d H 90 that is: S, Class O errors will be retained for 30 days, H errors will be retained for 90 days. You can use smit errclear to manually clear error logs.