Enter the root application information in message ranking prompt. On entering the root application information, from among the performance analysis traces that were read when message tracking started, the performance analysis trace information that matches with the entered root application information, is output chronologically. You can check places where response takes time or check traces for which the return code is in error, from the performance analysis trace information that is output.

(a) Output format

Execute prf command in message ranking prompt. For details on prf command, see fault analysis command in "cscmsgtk (Starting message tracking function) " in "Service Platform Reference Guide".

The execution format of prf command is as follows:

 
prf <IP address of root application information>/<Process ID of root application information>/<Communication number of root application information>
 

(b) Output format

The output format is as follows:

 
1
2
3
4
5

 
yyyy/MM/dd hh:mm:ss.SSS
No. interval(msec) event rc thread(hashcode) interface operation
No. + Time elapsed from the trace output before the last event ID return code thread ID interface name operation information
            Functionality layer: Trace collection point (corresponding event ID)
Delay level
 

The lines 3 to 5 output the number if traces for performance analysis.

Note that when the output mode is detailed output mode, the performance analysis trace which is collected from other than Service Platform is also output. For details on the output mode, see the output mode switching command in "cscmsgtk (Starting message tracking function) " in "Service Platform Reference Guide".

1^st Line: Start output Date and time

The date and time when performance analysis trace output started, is output in "yyyy/MM/dd hh:mm:ss.SSS" format.

• yyyy: Year
• MM: Month
• dd: Day
• hh: Hour
• mm: Minutes
• ss: Seconds
• SSS" Milliseconds

2^nd Line: Column

The columns corresponding to performance analysis trace are output. Contents same as the output format are output.

3^rd Line: Contents of Performance analysis trace

Outputs the contents of performance analysis trace corresponding to the columns.

Display item	Corresponding column	Description
No.	No.	Indicates a serial number.
Time elapsed from the trace output before the last	interval(msec)	Indicates the elapsed time (Milliseconds) from the performance analysis trace output before the earlier trace. Outputs single byte numeric characters 6 digits (Approximately 16 minutes). When the value exceeds this time, the "999999" is output.
Event ID	event	Indicates the Event ID for trace collection point.
Return code	rc	Indicates return code (hexadecimal). One of the following is displayed: 0: Normal Other than 0: Abnormal
Thread ID	thread(hashcode)	Indicates the thread ID of threads in process that has collected trace information, and hash value of threads. Thread ID Lower case hexadecimals with "0x" added. Hash value Lower case hexadecimals with "0x" added. If the value is less than 10 digits, digit alignment is performed to 10 digits. If the value is 10 digits or more, all digits are output. Note that if hexadecimal conversion fails, "???????? (????????)" is output. Also, the thread hash value might not be output for the trace information acquired using CTM and redirector. In such cases, only the thread ID is output.
Interface name	interface	Indicates the interface name for each collection point.
Operation information	operation	Indicates the operation information related to collection point.

4^th Line: Point name of performance analysis trace

The point name of performance analysis trace is output. "-" is output when performance analysis trace is collected from trace points other than Service Platform.

Display item	Description
Functionality layer	Indicates the layer of functionality that outputs performance analysis trace. One of the following is output: HCSC (MSG): Messaging infrastructure HCSC (BP): Business process infrastructure HCSC (DT): Data transformation infrastructure HCSC (FTR): FTP reception HCSC (HTR): HTTP reception HCSC (MQR): Message Queue reception HCSC (DBA): DB adapter HCSC (FFA): File adapter HCSC (OAA): Object Access adapter HCSC (MQA): Message Queue adapter HCSC (FTP): FTP adapter HCSC (FTI): FTP inbound adapter HCSC (FOA): File operations adapter HCSC (MA): Mail adapter HCSC (HTA): HTTP adapter
Trace collection point	Indicates the point that outputs performance analysis trace in each layer.
Corresponding event ID	Indicates the Event ID corresponding to acquired point. Outputs -corresponding-event-ID when the trace is acquired at starting point (inlet), and outputs corresponding-event-ID- when the trace is acquired at end point. (Outlet). Not output when there is no event ID.

5^th Line: Delay level

The delay level in trace is output.

Display items	Description
No line	Indicates delay level where elapsed milliseconds are within 1 second.
@@ (1sec-10sec)	Indicates delay level where elapsed milliseconds are greater than 1 second.
@@ @@ (10sec-100sec)	Indicates delay level where elapsed milliseconds are greater than 10 seconds.
@@ @@ @@ (100sec more...)	Indicates delay level where elapsed milliseconds are greater than 100 seconds.

(c) Output example

The example of output is as follows:

In the output example, there is a delay between the trace serial number 117 and trace serial number 121 and there is an error in return code of trace serial number 121.

Identify the performance analysis trace, whose information you want to check, from the return code or delay level displayed in "7.10.3(1) Displaying performance analysis trace that matches with root application information", and acquire the serial numbers of those performance analysis trace. If you enter the acquired serial number in continuation with the root application information, the detail information of that performance analysis trace is output. Outputs information such as message common ID, request ID, service component name and information useful for error analysis is acquired.

(a) Input format

Execute the prf command in message ranking prompt. Specify the serial number in continuation with root application information, for displaying detailed information of performance analysis trace. For details on the prf command, see fault analysis command in cscmsgtk (Starting message tracking function) " in "Service Platform Reference Guide".

The execution format of prf command is as follows:

 
prf <IP address of root application information>/<Process ID of root application information>/<Communication number of root application information > serial number
 

(b) Output format

The contents that are output to the ASCII column of performance analysis trace are output. Nothing is output when nothing is setup in the ASCII column.

(c) Output example

Output example is as follows:

• When detailed information of trace serial number 117 is output
  

  

• When detailed information of trace serial number 121 is output