settrace

The settrace command sets options to trace certain I/O operations through the system.

Syntax


>>- svctask -- -- settrace -- -- -type --+- mdisk -+-- --------->
                                         '- vdisk -'      

>-- -file -- filename -- -- -trigger --+- full ----+-- --------->
                                       +- status --+      
                                       +- command -+      
                                       +- timeout -+      
                                       +- trigger -+      
                                       '- abort ---'      

>--+----------+-- --+--------------+-- --+---------+-- --------->
   '- -abort -'     '- -timestamp -'     '- -data -'      

>--+--------+-- --+-----------+-- --+---------+-- -------------->
   '- -tag -'     '- -detect -'     '- -init -'      

>--+----------+-- --+---------+-- ------------------------------>
   '- -sense -'     '- -cmds -'      

>--+--------------------------+-- ------------------------------>
   '- -percent -- percentage -'      

>--+----------------------------------------------------+-- ---->
   '- -cmdlist -- cmd_list --+------------------------+-'      
                             '- -cmdmask -- cmd_mask -'        

>--+--------------------------------------------------------+--><
   '- -skcqlist -- skcq_list --+--------------------------+-'   
                               '- -skcqmask -- skcq_mask -'

Parameters

-type mdisk | vdisk

(Required) Specifies the type of objects to trace.

-file filename

(Required) Specifies the file name prefix for the trace file.

(Required) Specifies an action for when the trace is started (triggered).

full: Specifies to stop the trace when the trace buffer is full, for MDisks and VDisks.
status: Sets a trigger for when the specified SCSI status (-skcqlist) is reported in sense data, for MDisks and VDisks.
command: Specifies a trigger for when the given SCSI command (-cmdlist) is sent, for MDisks and VDisks.
timeout: Sets a trigger for when a timeout occurs, for MDisks only.
trigger: Specifies to keep running until the trigger event, for MDisks only.
abort: Sets a trigger for when an abnormal end occurs, for VDisks only.

-abort

(Optional) Adds abnormal ending details to the trace, for VDisks only.

-timestamp

(Optional) Adds a time-stamp to each entry in the trace. A file name is created from the prefix plus a time-stamp. The file name is in the form prefix_AAAAAA_YYMMDD_HHMMSS, where AAAAAA is the panel name of the node generating the trace file.

-data

(Optional) Adds I/O data to the trace.

-tag

(Optional) Adds CCB tags to the trace, for MDisks only.

-detect

(Optional) Adds MDisk discovery details to the trace, for MDisks only.

-init

(Optional) Adds MDisk initialization details to the trace, for MDisks only.

-sense

(Optional) Adds SCSI sense data to the trace, for VDisks only.

-cmds

(Optional) Adds commands data to the trace, for VDisks only.

-percent

(Optional) Specifies the trigger point in the trace file, which determines the amount of data to collect after the trigger point. The default value is 50, which places the trigger point in the middle of the trace file.

-cmdlist cmd_list

(Optional) Adds the commands in the cmd_list to the trace file.

-cmdmask cmd_mask

(Optional) Adds the commands in the cmd_mask to the trace file. The -cmdmask parameter must be used with the -cmdlist parameter.

-skcqlist skcq_list

(Optional) Specifies an SKCQ list, which adds only those SKCQ details to the trace file.

-skcqmask skcq_mask

(Optional) Specifies an SKCQ mask, which adds only those SKCQ details to the trace file. The -skcqmask parameter must be used with the -skcqlist parameter.

Description

The settrace command sets the various I/O tracing options for managed disks or virtual disks. When the relevant disk type trace is subsequently triggered, the options specify the data to be included in the trace file.

The file name specifies a file name prefix to use when you are generating a trace file. The system appends the node panel name and a timestamp to the file name.

A maximum of 10 trace files are kept on the cluster. When the eleventh trace is made, the oldest existing trace file is overwritten.

The directory can also hold files that are retrieved from other nodes. These files are not counted. The cluster deletes the oldest file to maintain the maximum number of files.

An invocation example

svctask settrace -type vdisk -file tracedump -trigger abort
 -percent 100 -abort -timestamp

The resulting output

No feedback