There are other traces you can perform to diagnose problems. This section discusses:
Occasionally, DB2 Customer Support may request that you take an administration tools trace. This trace is intended for DB2's graphical administrative tools (such as the Control Center, the Alert Center, and the Command Center), and provides additional information when these tools encounter a problem accessing or using a database. If you suspect the user interface itself to be the problem, take an ICLUI trace instead.
This trace is only recommended for problems that can be re-created. It provides detailed information on requests, connections, attachments to instances, CLI/ODBC/JDBC calls, DAS requests, API calls, and other activities that were happening when the problem occurs.
The format of the trace file is similar to the log file, except that more information is recorded. See Using the Administrative Tools Log, for more information.
Depending on how much information is written to it, the trace file can grow to the maximum file size imposed by the operating system. The file is overwritten the next time you perform an administration tools trace.
To invoke the administration tools trace facility:
set DBTTRACE=Y
set DBTTRACE_FILE=location
For example, the following command will send trace output to the file d:\aetrace:
set DBTTRACE_FILE=d:\aetrace
The DB2 Connect trace (ddcstrc) is useful for diagnosing problems with the DB2 Connect product. For more information, see the DB2 Connect User's Guide.
Use the DRDA trace (db2drdat command) to capture the DRDA data stream exchanged between a DRDA Application Requestor (AR) and the DRDA Application Server (AS). This tool is most often used for problem determination, but you can also use it for performance tuning in a client/server environment by determining how many sends and receives are required to execute an application. You need one of SYSADM, SYSCTRL, or SYSMAINT authority to use it.
The syntax for the db2drdat command is shown in the following diagram. Explanations of the parameters follow.
.---------.
.-on--. V |
>>-db2drdat--+-+-----+-----+--r-+--+---+-----------+-+---------><
| +--s-+ '--l=length-' |
| '--c-' |
'-off--+--------------+---+--------+----'
'--t=tracefile-' '--p=pid-'
Parameters
Logged CPI-C error information includes:
| Note: | If tracefile is not specified, messages are directed to the db2drdat.dmp file in the current directory. |
Notes:
DB2 generates alerts to notify database administrators or network administrators of a serious database error or database server error. It uses two standards of alerts:
When an SNA alert is generated, it includes different kinds of information about the problem that occurred. You can use this information to help correct whatever is causing a problem with SNA and APPC communications.
For information on taking an SNA trace, use the keywords "SNA trace" to search the DB2 Product and Service Technical Library at: http://www.ibm.com/software/data/db2/library/. The search produces product documentation. Select the platform that is appropriate to your environment. Check the diagnosis, troubleshooting, or problem determination information and look up "tracing" or "SNA trace" from the index.
Applications using CLI, ODBC, or SQLJ and the DB2 CLI Driver can have all the function calls traced to a plain text file. This can help with problem determination, database and application tuning or just to better understand what a third party application is doing.
The trace can be enabled at runtime using the CLI/ODBC Settings notebook, accessible from the Client Configuration Assistant if it is available, or by editing the db2cli.ini file directly. A CLI/ODBC/JDBC application can also enable the trace by setting the SQL_ATTR_TRACE and SQL_ATTR_TRACEFILE Environment or Connection attributes. These are the same attributes used by the tracing facility in the Microsoft ODBC Driver Manager.
The db2cli.ini file is located by default in the \sqllib\ path for Intel platforms, and the /sqllib/cfg/ path for UNIX platforms. The CLI/ODBC/JDBC configuration keywords used by the trace facility are:
The following lines must be added to enable the trace. Keywords are NOT case-sensitive.
or
TracePathname=(fully qualified pathname)
| Note: | At least one of these two must be specified, but not both. |
For example:
[COMMON] trace=1 tracefilename=d:\temp\clitrace.txt
Setting TRACE to 0 turns tracing off. The trace file information can be left in the configuration file for the next time it is needed. See TRACE for more information.
If the application does not exit or exits abnormally, the trace file will probably not be complete. Setting TRACEFLUSH to 1 will cause a flush to disk on every function call (which will increase the overhead of tracing dramatically). See TRACEFLUSH for more information.
To have information about each network request included in the trace file, set TRACECOMM to 1. See TRACECOMM for more information.
If you want the timing information added to each line in the CLI trace, set TRACETIMESTAMP to 1, 2, or 3. Setting TRACETIMESTAMP to 1 provides the absolute time in seconds and milliseconds, followed by the timestamp. Setting TRACETIMESTAMP to 2 provides the absolute time in seconds and milliseconds. Setting TRACETIMESTAMP to 3 provides the timestamp.
If you want the process and thread identifier added to each line in the CLI trace, set TRACEPIDTID to 1.
The db2cli.ini file is located by default in the \sqllib\ path for Intel platforms, and the /sqllib/cfg/ path for UNIX platforms. Use the following CLI/ODBC/JDBC configuration keywords to enable a JDBC trace which provides more information than a regular CLI/ODBC trace. The CLI/ODBC/JDBC configuration keywords used by the trace facility are:
The following lines must be added to enable the trace. Keywords are NOT case-sensitive.
JDBCPathname=(fully qualified pathname)
For example:
[COMMON] jdbctrace=1 jdbcpathname=d:\temp
Setting JDBCTRACE to 0 turns tracing off. The trace file information can be left in the configuration file for the next time it is needed. See JDBCTRACE for more information.
If the application does not exit or exits abnormally, the trace file will probably not be complete. Setting JDBCTRACEFLUSH to 1 will cause a flush to disk on every function call (which will increase the overhead of tracing dramatically). See JDBCTRACEFLUSH for more information.
If you used a fully qualified filename with the TRACEFILENAME keyword, you should have no problem locating the file. If you used a relative pathname, it will depend on what the operating system considers the current path of the application.
If you used a pathname instead of a filename with the TRACEPATHNAME or JDBCTRACEPATHNAME keyword, you will need to check the directory for a set of files created with the name set to the process id of the application and an extension that is a sequence number for each unique thread (for example, 65397.0, 65397.1, 65397.2). The file date and timestamp can be used to help locate the relevant file.
If you used a relative pathname, it will depend on what the operating system considers the current path of the application.
If there is no output file:
The purpose of the trace is to display the sequence of calls, the input and output arguments and the return code for each function called. The trace is intended for people familiar with the CLI/ODBC/JDBC function calls. Two important items, are the SQL statement text being executed, and any error messages that the application may not be reporting.
To locate:
Search the trace file for the strings "SQLExecDirect" and "SQLPrepare", you will find the SQL Statement on the same line that contains the text and the "--->" input arrow (your editor may wrap the line).
Search the trace file for "SQLError", the message text will be shown on the line that contains the string and the output arrow "<---".
Also, search for SQLGetDiagRec() or SQLGetDiagField() following the receipt of SQL_ERROR or SQL_SUCCESS_WITH_INFO return codes.
Search for "Unretrieved error message=" This indicates that a previous call got an SQL_ERROR or SQL_SUCCESS_WITH_INFO return code, but that the application did not query for the error information.
| Note: | An application may expect some error messages, you should look at all the error messages in the trace file and try to determine the serious ones. |
Refer to the example trace file below. Note that the line numbers have been added for this discussion, and do NOT appear in the trace.
| Note: | the granularity or accuracy of these timings vary between platforms. |
This example has line numbers added to aid the discussion, line numbers do NOT appear in the trace.
1 Build Date: 97/05/13-Product: QDB2/6000 (4) - Driver Version: 05.00.0000
2 SQLAllocHandle( fHandleType=SQL_HANDLE_ENV, hInput=0:0,
phOutput=&2ff7f388 )
3 ---> Time elapsed - +1.399700E-002 seconds
4 SQLAllocHandle( phOutput=0:1 )
5 <--- SQL_SUCCESS Time elapsed - +6.590000E-003 seconds
6 SQLAllocHandle( fHandleType=SQL_HANDLE_DBC, hInput=0:1,
phOutput=&2ff7f378 )
7 ---> Time elapsed - +1.120000E-002 seconds
8 SQLAllocHandle( phOutput=0:1 )
9 <--- SQL_SUCCESS Time elapsed - +8.979000E-003 seconds
10 SQLSetConnectOption( hDbc=0:1, fOption=SQL_ATTR_AUTOCOMMIT, vParam=0 )
11 ---> Time elapsed - +6.638000E-003 seconds
12 SQLSetConnectOption( )
13 <--- SQL_SUCCESS Time elapsed - +1.209000E-003 seconds
14 SQLDriverConnect( hDbc=0:1, hwnd=0:0, szConnStrIn="DSN=loopback;
uid=clitest1;pwd=*******", cbConnStrIn=-3, szConnStrOut=&2ff7e7b4,
cbConnStrOutMax=250, pcbConnStrOut=&2ff7e7ae,
fDriverCompletion=SQL_DRIVER_NOPROMPT )
15 ---> Time elapsed - +1.382000E-003 seconds
16 SQLDriverConnect( szConnStrOut="DSN=LOOPBACK;UID=clitest1;PWD=*******;",
pcbConnStrOut=38 )
17 <--- SQL_SUCCESS Time elapsed - +7.675910E-001 seconds
18 ( DSN="LOOPBACK" )
19 ( UID="clitest1" )
20 ( PWD="*******" )
21 SQLAllocHandle( fHandleType=SQL_HANDLE_STMT, hInput=0:1,
phOutput=&2ff7f378 )
22 ---> Time elapsed - +1.459900E-002 seconds
23 SQLAllocHandle( phOutput=1:1 )
24 <--- SQL_SUCCESS Time elapsed - +7.008300E-002 seconds
25 SQLExecDirect( hStmt=1:1, pszSqlStr="create table test(id integer,
name char(20), created date)", cbSqlStr=-3 )
26 ---> Time elapsed - +1.576899E-002 seconds
27 SQLExecDirect( )
28 <--- SQL_SUCCESS Time elapsed - +1.017835E+000 seconds
29 SQLPrepare( hStmt=1:1, pszSqlStr="insert into test
values (?, ?, current date)", cbSqlStr=-3 )
30 ---> Time elapsed - +5.008000E-003 seconds
31 SQLPrepare( )
32 <--- SQL_SUCCESS Time elapsed - +7.896000E-003 seconds
33 SQLBindParameter( hStmt=1:1, iPar=1, fParamType=SQL_PARAM_INPUT,
fCType=SQL_C_LONG, fSQLType=SQL_INTEGER, cbColDef=4, ibScale=0,
rgbValue=&20714d88, cbValueMax=4, pcbValue=&20714d54 )
34 ---> Time elapsed - +2.870000E-003 seconds
35 SQLBindParameter( )
36 <--- SQL_SUCCESS Time elapsed - +3.803000E-003 seconds
37 SQLBindParameter( hStmt=1:1, iPar=2, fParamType=SQL_PARAM_INPUT,
fCType=SQL_C_CHAR, fSQLType=SQL_CHAR, cbColDef=20, ibScale=0,
rgbValue=&20714dd8, cbValueMax=21, pcbValue=&20714da4 )
38 ---> Time elapsed - +2.649000E-003 seconds
39 SQLBindParameter( )
40 <--- SQL_SUCCESS Time elapsed - +3.882000E-003 seconds
41 SQLExecute( hStmt=1:1 )
42 ---> Time elapsed - +3.681000E-003 seconds
43 ( iPar=1, fCType=SQL_C_LONG, rgbValue=10, pcbValue=4, piIndicatorPtr=4 )
44 ( iPar=2, fCType=SQL_C_CHAR, rgbValue="-3", pcbValue=2, piIndicatorPtr=2 )
45 SQLExecute( )
46 <--- SQL_SUCCESS Time elapsed - +4.273490E-001 seconds
47 SQLExecute( hStmt=1:1 )
48 ---> Time elapsed - +5.483000E-003 seconds
49 ( iPar=1, fCType=SQL_C_LONG, rgbValue=10, pcbValue=4, piIndicatorPtr=4 )
50 ( iPar=2, fCType=SQL_C_CHAR, rgbValue="-3", pcbValue=2, piIndicatorPtr=2 )
51 SQLExecute( )
52 <--- SQL_SUCCESS Time elapsed - +1.299300E-002 seconds
53 SQLExecute( hStmt=1:1 )
54 ---> Time elapsed - +3.702000E-003 seconds
55 ( iPar=1, fCType=SQL_C_LONG, rgbValue=10, pcbValue=4, piIndicatorPtr=4 )
56 ( iPar=2, fCType=SQL_C_CHAR, rgbValue="-3", pcbValue=2, piIndicatorPtr=2 )
57 SQLExecute( )
58 <--- SQL_SUCCESS Time elapsed - +1.265700E-002 seconds
59 SQLExecDirect( hStmt=1:1, pszSqlStr="select * from test", cbSqlStr=-3 )
60 ---> Time elapsed - +2.983000E-003 seconds
61 SQLExecDirect( )
62 <--- SQL_SUCCESS Time elapsed - +2.469180E-001 seconds
63 SQLBindCol( hStmt=1:1, iCol=1, fCType=SQL_C_LONG, rgbValue=&20714e38,
cbValueMax=4, pcbValue=&20714e04 )
64 ---> Time elapsed - +5.069000E-003 seconds
65 SQLBindCol( )
66 <--- SQL_SUCCESS Time elapsed - +2.660000E-003 seconds
67 SQLBindCol( hStmt=1:1, iCol=2, fCType=SQL_C_CHAR, rgbValue=&20714e88,
cbValueMax=21, pcbValue=&20714e54 )
68 ---> Time elapsed - +2.492000E-003 seconds
69 SQLBindCol( )
70 <--- SQL_SUCCESS Time elapsed - +2.795000E-003 seconds
71 SQLBindCol( hStmt=1:1, iCol=3, fCType=SQL_C_CHAR, rgbValue=&20714ee8,
cbValueMax=21, pcbValue=&20714eb4 )
72 ---> Time elapsed - +2.490000E-003 seconds
73 SQLBindCol( )
74 <--- SQL_SUCCESS Time elapsed - +2.749000E-003 seconds
75 SQLFetch( hStmt=1:1 )
76 ---> Time elapsed - +2.660000E-003 seconds
77 SQLFetch( )
78 <--- SQL_SUCCESS Time elapsed - +9.200000E-003 seconds
79 ( iCol=1, fCType=SQL_C_LONG, rgbValue=10, pcbValue=4 )
80 ( iCol=2, fCType=SQL_C_CHAR, rgbValue="-3 ",
pcbValue=20 )
81 ( iCol=3, fCType=SQL_C_CHAR, rgbValue="1997-05-23", pcbValue=10 )
82 SQLFetch( hStmt=1:1 )
83 ---> Time elapsed - +4.942000E-003 seconds
84 SQLFetch( )
85 <--- SQL_SUCCESS Time elapsed - +7.860000E-003 seconds
86 ( iCol=1, fCType=SQL_C_LONG, rgbValue=10, pcbValue=4 )
87 ( iCol=2, fCType=SQL_C_CHAR, rgbValue="-3 ",
pcbValue=20 )
88 ( iCol=3, fCType=SQL_C_CHAR, rgbValue="1997-05-23", pcbValue=10 )
89 SQLFetch( hStmt=1:1 )
90 ---> Time elapsed - +4.872000E-003 seconds
91 SQLFetch( )
92 <--- SQL_SUCCESS Time elapsed - +7.669000E-003 seconds
93 ( iCol=1, fCType=SQL_C_LONG, rgbValue=10, pcbValue=4 )
94 ( iCol=2, fCType=SQL_C_CHAR, rgbValue="-3 ",
pcbValue=20 )
95 ( iCol=3, fCType=SQL_C_CHAR, rgbValue="1997-05-23", pcbValue=10 )
96 SQLFetch( hStmt=1:1 )
97 ---> Time elapsed - +5.103000E-003 seconds
98 SQLFetch( )
99 <--- SQL_NO_DATA_FOUND Time elapsed - +6.044000E-003 seconds
100 SQLCloseCursor( hStmt=1:1 )
101 ---> Time elapsed - +2.682000E-003 seconds
102 SQLCloseCursor( )
103 <--- SQL_SUCCESS Time elapsed - +6.794000E-003 seconds
104 SQLExecDirect( hStmt=1:1, pszSqlStr="select * foo bad sql", cbSqlStr=-3 )
105 ---> Time elapsed - +2.967000E-003 seconds
106 SQLExecDirect( )
107 <--- SQL_ERROR Time elapsed - +1.103700E-001 seconds
108 SQLError( hEnv=0:0, hDbc=0:0, hStmt=1:1, pszSqlState=&2ff6f19c,
pfNativeError=&2ff6ed00, pszErrorMsg=&2ff6ed9c, cbErrorMsgMax=1024,
pcbErrorMsg=&2ff6ed0a )
109 ---> Time elapsed - +2.267000E-003 seconds
110 SQLError( pszSqlState="42601", pfNativeError=-104,
pszErrorMsg="[IBM][CLI Driver][DB2/6000] SQL0104N An unexpected
token "foo bad sql" was found following "select * ".
Expected tokens may include: "<space>". SQLSTATE=42601
111 ", pcbErrorMsg=163 )
112 <--- SQL_SUCCESS Time elapsed - +5.299000E-003 seconds
113 SQLError( hEnv=0:0, hDbc=0:0, hStmt=1:1, pszSqlState=&2ff6f19c,
pfNativeError=&2ff6ed00, pszErrorMsg=&2ff6ed9c, cbErrorMsgMax=1024,
pcbErrorMsg=&2ff6ed0a )
114 ---> Time elapsed - +2.753000E-003 seconds
115 SQLError( )
116 <--- SQL_NO_DATA_FOUND Time elapsed - +2.502000E-003 seconds
117 SQLExecDirect( hStmt=1:1, pszSqlStr="select * foo bad sql", cbSqlStr=-3 )
118 ---> Time elapsed - +3.292000E-003 seconds
119 SQLExecDirect( )
120 <--- SQL_ERROR Time elapsed - +6.012500E-002 seconds
121 SQLFreeHandle( fHandleType=SQL_HANDLE_STMT, hHandle=1:1 )
122 ---> Time elapsed - +2.867000E-003 seconds
123 ( Unretrieved error message="SQL0104N An unexpected token "foo bad sql"
was found following "select * ". Expected tokens may
include: "<space>". SQLSTATE=42601
124 " )
125 SQLFreeHandle( )
126 <--- SQL_SUCCESS Time elapsed - +4.936600E-002 seconds
127 SQLEndTran( fHandleType=SQL_HANDLE_DBC, hHandle=0:1, fType=SQL_ROLLBACK )
128 ---> Time elapsed - +2.968000E-003 seconds
129 SQLEndTran( )
130 <--- SQL_SUCCESS Time elapsed - +1.643370E-001 seconds
131 SQLDisconnect( hDbc=0:1 )
132 ---> Time elapsed - +2.559000E-003 seconds
133 SQLDisconnect( )
134 <--- SQL_SUCCESS Time elapsed - +8.253310E-001 seconds
135 SQLFreeHandle( fHandleType=SQL_HANDLE_DBC, hHandle=0:1 )
136 ---> Time elapsed - +4.247000E-003 seconds
137 SQLFreeHandle( )
138 <--- SQL_SUCCESS Time elapsed - +4.742000E-003 seconds
139 SQLFreeHandle( fHandleType=SQL_HANDLE_ENV, hHandle=0:1 )
140 ---> Time elapsed - +2.023000E-003 seconds
141 SQLFreeHandle( )
142 <--- SQL_SUCCESS Time elapsed - +4.420000E-003 seconds
For the trace to be of any use for multi-threaded or multi-process applications, you will need to use the TRACEPATHNAME keyword. Otherwise the trace will be garbled if multiple threads or processes are writing to it simultaneously. See TRACEPATHNAME for more information.
The files are created in the path specified with the name set to the process id of the application and an extension that is a sequence number for each unique thread (for example, 65397.0, 65397.1, 65397.2).
By having each thread write to its own file, no semaphores are needed to control access to the tracefile, which means tracing doesn't change the behavior of a multi-thread application.
It is useful to understand the difference between the ODBC trace provided by the ODBC Driver Manager and the DB2 CLI/ODBC driver (IBM ODBC Driver Tracing).
Output file formats are different. The distinction is that the ODBC trace will show the calls made by the application to the Driver Manager. The DB2 CLI trace shows the calls received from the ODBC Driver manager.
The ODBC driver manager may map application function calls to either different functions, different arguments or may delay the call.
One or more of the following may apply:
For these reasons you may need to enable and compare the output of both traces to get a clear picture of what is happening.
For more information, refer to the Microsoft ODBC 3.0 Software Development Kit and Programmer's Reference.
The CLI/ODBC/JDBC configuration keywords used by the trace facility are:
When this option is on (1), JDBC trace records are appended to files in the subdirectory indicated by the JDBCTRACEPATHNAME configuration parameter.
For example, to set up a JDBC trace file that is written to disk after each trace entry:
[COMMON]
JDBCTRACE=1
JDBCTRACEPATHNAME=E:\JDBCTRACES\CLI
JDBCTRACEFLUSH=1
(This option is contained in the Common section of the initialization file and therefore applies to all connections to DB2.)
Set this option on (JDBCTRACEFLUSH = 1) to force a write to disk after each trace entry. This will slow down the trace process, but will ensure that each entry is written to disk before the application continues to the next statement.
This option is only used when the JDBCTRACE CLI/ODBC/JDBC option is turned on. See the JDBC TRACE option for an example.
This option is contained in the Common section of the initialization file and therefore applies to all connections to DB2.
Each thread or process that uses the same DLL or shared library will have a separate JDBC trace file created in the specified directory.
No trace will occur, and no error message will be returned, if the subdirectory given is invalid or if it cannot be written to.
This option is only used when the JDBCTRACE option is turned on. This will be done automatically when you set this option in the CLI/ODBC/JDBC Configuration utility.
See the JDBCTRACE option for an example of using the various trace settings.
JDBC trace should only be used for debugging purposes. It will slow down the execution of the JDBC driver, and the trace information can grow quite large if it is left on for extended periods of time.
This option is contained in the Common section of the initialization file and therefore applies to all connections to DB2.
When this option is on (1), CLI/ODBC/JDBC trace records are appended to the file indicated by the TRACEFILENAME configuration parameter or to files in the subdirectory indicated by the TRACEPATHNAME configuration parameter.
For example, to set up a CLI/ODBC/JDBC trace file that is written to disk after each trace entry:
[COMMON]
TRACE=1
TRACEFILENAME=E:\TRACES\CLI\MONDAY.CLI
TRACEFLUSH=1
(This option is contained in the Common section of the initialization file and therefore applies to all connections to DB2.)
When TRACECOMM is set on (1) then information about each network request will be included in the trace file.
This option is only used when the TRACE CLI/ODBC/JDBC option is turned on. See the TRACE option for an example.
This option is contained in the Common section of the initialization file and therefore applies to all connections to DB2.
If the file specified does not exist, then it will be created; otherwise, the new trace information will be appended to the end of the file.
If the filename given is invalid or if the file cannot be created or written to, no trace will occur and no error message will be returned.
This option is only used when the TRACE option is turned on. This will be done automatically when you set this option in the CLI/ODBC/JDBC Configuration utility.
See the TRACE option for an example of using the various trace settings. The TRACEPATHNAME option will be ignored if this option is set. TRACEPATHNAME should be used instead of TRACEFILENAME for multi-threaded applications so that the thread information is kept separate.
CLI/ODBC/JDBC trace should only be used for debugging purposes. It will slow down the execution of the CLI/ODBC/JDBC driver, and the trace information can grow quite large if it is left on for extended periods of time.
(This option is contained in the Common section of the initialization file and therefore applies to all connections to DB2.)
Set this option on (TRACEFLUSH = 1) to force a write to disk after each trace entry. This will slow down the trace process, but will ensure that each entry is written to disk before the application continues to the next statement.
This option is only used when the TRACE CLI/ODBC/JDBC option is turned on. See the TRACE option for an example.
This option is contained in the Common section of the initialization file and therefore applies to all connections to DB2.
Each thread or process that uses the same DLL or shared library will have a separate CLI/ODBC/JDBC trace file created in the specified directory.
No trace will occur, and no error message will be returned, if the subdirectory given is invalid or if it cannot be written to.
This option is only used when the TRACE option is turned on. This will be done automatically when you set this option in the CLI/ODBC/JDBC Configuration utility.
See the TRACE option for an example of using the various trace settings. It will be ignored if the CLI/ODBC/JDBC option TRACEFILENAME is used.
CLI/ODBC/JDBC trace should only be used for debugging purposes. It will slow down the execution of the CLI/ODBC/JDBC driver, and the trace information can grow quite large if it is left on for extended periods of time.
This option is contained in the Common section of the initialization file and therefore applies to all connections to DB2.
This option is only used when the TRACE option is turned on. This will be done automatically when you set this option in the CLI/ODBC/JDBC Configuration utility.
If you want the timing information added to each line in the CLI trace, set TRACETIMESTAMP to 1, 2, or 3. Setting TRACETIMESTAMP to 1 provides the absolute time in seconds and milliseconds, followed by the timestamp. Setting TRACETIMESTAMP to 2 provides the absolute time in seconds and milliseconds. Setting TRACETIMESTAMP to 3 provides the timestamp.
This option is only used when the TRACE option is turned on. This will be done automatically when you set this option in the CLI/ODBC/JDBC Configuration utility.
If you want the process and thread identifier added to each line in the CLI trace, set TRACEPIDTID to 1.