15.2.2 Options
(1) -r {RDAREA-name[,RDAREA-name]...|ALL}...
~<identifier> ((1-30))
Specifies the names of the RDAREAs to be analyzed.
You can specify the following RDAREA types:
- Data dictionary RDAREA
- Data dictionary LOB RDAREA
- User RDAREA (temporary table RDAREAs are excluded from RDAREA logical analysis)
- User LOB RDAREA
- Registry RDAREA
- Registry LOB RDAREA
For details about specifying RDAREAs, see 1.5.2 Specification of RDAREAs in operation commands and utilities.
When ALL is specified, all the above RDAREAs are subject to processing. Any other RDAREAs are skipped.
- Rules
- If you are also specifying the -q option, specify the original RDAREA name.
- The condition analysis result accumulation facility and the facility for predicting reorganization time use a single transaction to analyze all tables stored in an RDAREA. Therefore, the value of the pd_max_access_tables operand in the system definition must at least equal the number of tables in the RDAREA that contains the greatest number of tables among all target RDAREAs.
(2) -k {logi|phys|clus|pred}
Specifies the type of analysis that is to be executed (by the database condition analysis facility, condition analysis result accumulation facility, or facility for predicting reorganization time).
- logi:
- Specifies that RDAREA condition analysis (logical analysis) or the condition analysis result accumulation facility is to be executed.
- phys:
- Specifies that RDAREA condition analysis (physical analysis) is to be executed.
- clus:
- Specifies that storage condition analysis is to be executed on cluster keys and clustering data pages.
- pred:
- Specifies that the facility for predicting reorganization time is to be executed.
- Combination of -k option and other options:
- The values of the -r, -t, and -i options depend on the value of the -k option:
-k option value | Whether or not operation can be specified | Remarks |
---|
-r | -t | -i |
---|
logi | M | N | N | Condition analysis by RDAREA (logical analysis) |
phys | M | N | N | Condition analysis by RDAREA (physical analysis) |
clus | N | M | N | Storage condition analysis on cluster keys and clustering data pages |
Omitted | N | M | N | Condition analysis by table |
N | N | M | Condition analysis by index |
M | N | N | Condition analysis by RDAREA (logical analysis) |
pred | O | N | N | Facility for predicting reorganization time |
- Legend:
- M: Specification is mandatory
- O: Specification is optional
- N: Specification is not permitted
(3) -d
Specifies that detailed page information is to be displayed. This option is applicable to logical analysis (-k logi) and condition analysis by table and by index. Whether or not pdrorg and pdreclaim need to be executed depends on this detailed page information.
(4) -z
Specifies that information about rows for tables and indexes stored in the page is to be displayed. Based on the information displayed by this option, determine whether managed rows in deleted status are to be released by executing the free page release utility (pdreclaim). The following table shows the information that can be displayed for each analysis target.
No. | Information that is displayed | RDAREA logical analysis | RDAREA physical analysis | Analysis by table | Index analysis | Cluster key analysis |
---|
1 | Table information | Number of managed rows in deleted status# | Y | N | Y | N | N |
2 | Index information | Number of rows managed by remaining entries | Y | N | N | Y | N |
3 | Total number of rows managed by indexes | Y | N | N | Y | N |
- Legend:
- Y: Can be displayed.
- N: Cannot be displayed.
- #
- Displays the total number of the following rows:
- Rows being deleted (waiting for completion of transaction)
- Rows that have been deleted (transaction completed)
- Rules
- When you specify this option, you must also specify the -d option.
(5) -f
Specifies that HiRDB file information is to be displayed. This option is applicable to physical analysis (-k phys) only.
(6) -a
Specifies that the analysis results are to be displayed in DAT format.
When this option is specified for RDAREA condition analysis (physical analysis), the utility assumes that the -f option is specified.
(7) -h
When the -a option is specified, specifies that a header is to be displayed.
(8) -u authorization-identifier
Specifies the authorization identifier of the user executing pddbst.
- Rules
- If this option is omitted, the authorization identifier and password defined in PDUSER in the client environment definition are assumed. If the PDUSER value is not set, the user name corresponding to the OS user ID of the user executing the utility is assumed.
- If you enclose an authorization identifier between double quotation marks ("), the system treats it as being case sensitive; otherwise, the system treats it as in all uppercase letters. If you are using sh (Bourne shell), csh (C shell), or ksh (Korn shell), you must enclose the entire user ID in single quotation marks (').
(9) -p password
Specifies the password for the authorization identifier specified in the -u option.
- Rules
- If this option is not specified and PDUSER is also not set in the client environment definition, the value entered on the utility execution screen is assumed.
- If you enclose a password between double quotation marks ("), the system treats it as being case sensitive; otherwise, the system treats it as in all uppercase letters. If you are using sh (Bourne shell), csh (C shell), or ksh (Korn shell), you must enclose the entire password in single quotation marks (').
(10) -t [authorization-identifier.]table-identifier
Specifies the name of the table to be analyzed.
- Rules
- You cannot specify a view table or a temporary table. If pddbst is executed with a view table or temporary table specified, the KFPK10106-W message is issued and pddbst terminates.
- If the authorization identifier is omitted, the authorization identifier specified in the -u option is assumed.
- If you enclose an authorization identifier or table identifier between double quotation marks ("), the system treats it as being case sensitive; otherwise, the system treats it as in all uppercase letters.
- If the table identifier contains a space, enclose it between double quotation marks ("). If you are using sh (Bourne shell), csh (C shell), or ksh (Korn shell), you must enclose the entire identifier in single quotation marks (').
(11) -s
Specifies that the display in the case of table condition analysis is to include the total number of lines stored in the table or the number of lines stored for each RDAREA.
When you run this utility concurrently with an online application, also specify -b. If -b is not specified in such a case, performance delays might occur in the online application because pages that have been read by the application are removed from the global buffer when pddbst reads and references the database.
(12) -i [authorization-identifier.]index-identifier
Specifies the name of the index to be analyzed.
- Rules
- You cannot specify a temporary table index. If pddbst is executed with a temporary table index specified, the KFPK10106-W message is issued and pddbst terminates.
- If the authorization identifier is omitted, the authorization identifier specified in the -u option is assumed.
- If you enclose an authorization identifier or index identifier between double quotation marks ("), the system treats it as being case sensitive; otherwise, the system treats it as in all uppercase letters.
- If the index identifier contains a space, enclose the entire index identifier between double quotation marks ("). If you are using sh (Bourne shell), csh (C shell), and ksh (Korn shell), you need to enclose the entire identifier in single quotation marks (').
(13) -q generation-number
~<unsigned integer> ((0-10))
Specifies that information about the inner replica facility is to be displayed.
- Rules
- An error results if this option is specified when HiRDB Staticizer Option has not been installed.
- When this option is specified, the utility analyzes the RDAREAs that contain the specified replica definition generation. If there are no RDAREAs subject to analysis in the specified generation, an error (KFPK10102-W message) results.
- When this option is omitted, the utility assumes that the following RDAREAs are to be analyzed:
When ALL is specified in the -r option to perform condition analysis by RDAREA: RDAREAs that can be specified as described in 15.2.2(1) -r {RDAREA-name[,RDAREA-name]...|ALL}...
When batch or direct specification is specified in the -r option to perform condition analysis by RDAREA: RDAREAs that are specified in the -r option
For condition analysis by table or index or storage condition analysis on cluster key and clustering data page: Current RDAREA
- This option cannot be specified when you execute the condition analysis result accumulation facility or the facility for predicting reorganization time. When the inner replica facility is used, the utility assumes -q 0.
(14) -b
Specifies that the online performance of applications is to be protected from any adverse effects of executing pddbst.
When the -s option is omitted, the utility assumes that this option is specified.
When this option is specified together with the -s option, the analysis time may increase. The -b option is ignored when searching data dictionary tables for a table and indexes that are to be the targets of database condition analysis, and when searching the operation management table by the facility for predicting reorganization time.
- Criteria
- pddbst first reads a resource page to be analyzed into the global buffer. If a space shortage occurs in the global buffer at that time, online performance may be affected adversely because a page that had already been read by an application may be swept out of the global buffer. To avoid such adverse effects on performance, specify the -b option.
- When the -b option is specified, a page read by pddbst is always treated as the oldest page, so that it will be the first page swept out of the global buffer in the event of a space shortage. When the -b option is not specified, the LRU management method is used.
- Advantages
- Even if a space shortage occurs in the global buffer, there will be no adverse effects on the online performance of applications. However, during execution of pddbst, re-reading of directory pages may occur in order to acquire the requisite number of rows.
(15) -e prediction-level
Specifies the prediction level to be used when the condition analysis result accumulation facility or the facility for predicting reorganization time is applied.
- 1
- Specifies that the condition analysis result accumulation facility or the facility for predicting reorganization time is to be executed in prediction level 1.
- 2
- Specifies that the condition analysis result accumulation facility or the facility for predicting reorganization time is to be executed in prediction level 2.
- Rules
- Set the same prediction level for both the condition analysis result accumulation facility and the facility for predicting reorganization time.
- When the -e option is specified, the utility assumes that the -b option is also specified.
(16) -w pause-time,segments-count
Specifies that the condition analysis result accumulation facility in prediction level 2 is to be executed with the interval analysis method. For details about interval analysis, see 15.6.1 Interval analysis. This method pauses processing for the specified amount of time after the specified number of segments have been analyzed.
- pause-time
- ~<unsigned integer> ((10-60000))
- Specifies the pause time, in units of 10 milliseconds. If the specified value is not an even 10 milliseconds, the value is rounded up (for example, 15 is rounded up to 20).
- segments-count ~<unsigned integer> ((1-2147483647))
- Specifies the number of segments to be analyzed between pauses.
- Rules
- The -w option is applicable to prediction level 2 (-e 2). Specifying this option in the case of another prediction level results in a control statement error.
- In the case of a HiRDB parallel server configuration, the specified value applies to all servers.
- If the global buffer contains no page to be analyzed, as many input/output operations as (specified number of segments x segment size) occur until the pause time is reached.
(17) -n analysis-segments-count
~<unsigned integer> ((2-10))
Specifies that the condition analysis result accumulation facility in prediction level 2 is to be executed with the merge analysis method. For details about merge analysis, see 15.6.2 Merge analysis.
This method divides a single execution of the condition analysis result accumulation facility into the specified number of analysis segments. Data equivalent to 1/n, where n is the specified number of analysis segments, is the target of each analysis. This means that it takes n executions to analyze all of the target data.
- Rules
- The -n option is applicable to prediction level 2 (-e 2). Specifying this option in the case of another prediction level results in a control statement error.
- When you specify the -n option, you must execute pddbst the specified number of times. If you change the -n option value before the current series of analyses has been completed, the utility ignores the analysis information that has been accumulated so far and starts the analyses over from the beginning.
- If maintenance is performed on the target table or index before the entire series of analyses has been completed, the utility starts the analyses over from the beginning.
(18) -I
(19) -m
Specifies that information about both the scheduled database maintenance date and the maintenance method is to be displayed by the facility for predicting reorganization time. When this option is omitted, the utility displays information about the scheduled database maintenance date only.
(20) -R monitoring-interval[,maintenance-extension-period]
When the facility for predicting reorganization time is executed, specifies the target period and the maintenance extension period.
- monitoring-interval ~<unsigned integer> ((1-400)) <<14>>
- Specifies the target monitoring interval in days.
- For example, to monitor an RDAREA for the next two weeks for scheduling a database maintenance date, specify 14.
- maintenance-extension-period ~<unsigned integer> ((0-400)) <<monitoring period/2>>
- Specifies the minimum number of days that must elapse between scheduled database maintenance dates.
- For example, once database maintenance has been executed, if you wish to require at least one week to elapse before the next database maintenance, specify 7.
(21) -c standard-value-definition-file-name
~<path name> ((up to 1,023 bytes))
When the facility for predicting reorganization time is executed, the utility makes its predictions based on standard values predefined for various items. To change any of the standard values, this option specifies the absolute or relative path name of the standard value definition file.
You specify in the standard value definition file the new standard values for the items. For details about the standard value definition file, see 15.5 Standard value definition file (facility for predicting reorganization time).
(22) -X response-monitoring-time-for-server-to-server-communication
~<unsigned integer> ((1-65535)) ((300))
If an error, such as a communication error, occurs at the server where the command was executed, the command may stop responding and the application may stop. To help you detect errors, pddbst enables you to monitor a response time during communication for dictionary manipulation that is performed by the command.
You set in the -X option a response monitoring time during dictionary manipulation (in seconds). If the execution time during dictionary manipulation exceeds the value set in the -X option, pddbst treats it as a dictionary access error and cancels processing with return code 8.
- Criteria
- If you want to detect an error in less time than 300 seconds in the event of a no-response from the server due to a communication error or unit down, specify a value less than 300 in the -X option.
- If the system switchover facility is used, the command may keep waiting for a response even though system switchover has been completed. In such a case, you can terminate the command immediately by reducing the monitoring time.
- The specified monitoring time may result in a timeout if a response from the dictionary is delayed and the utility's preprocessing is not completed within 300 seconds, which is the default value of the -X option. This can happen when many applications and utilities are executing concurrently. In such an environment, specify a value greater than 300 in the -X option.
(23) -v control-statement-file-name
~<path name> ((up to 1,023 bytes))
Specifies the absolute or relative path name of the file that contains the pddbst control statements.
This file must be located on the host where pddbst is executed. You can specify the following control statements; note that comments cannot be specified in the control statements file.
- predict statement
- workdir statement
- option statement
(24) -W execution-monitoring-interval
~<unsigned integer> ((0 to 3600))
Specifies (in minutes) the monitoring interval when the execution time of the database condition analysis utility is to be monitored.
For guidelines on the value to specify and details about the resulting operation, see the description of the pd_cmd_exec_time operand in the system common definition in the manual HiRDB Version 9 System Definition.
If 0 is specified in this option, the utility's execution time is not monitored.
If this option is omitted, the value of the pd_cmd_exec_time operand in the system common definition takes effect.