• qufile001="import-information-queue-filename"...qufile008="import-information-queue-filename"
  ~ <[pathname/]filename of 1-64 bytes>
  Specify the names of the import information queue files, as absolute or relative pathnames. If you specify a relative pathname, Datareplicator assumes $HDSPATH/relative-pathname as the absolute pathname. In this case, specify the pathname so that the final absolute pathname will not exceed 125 characters.
  Datareplicator uses file swapping to store information in the import information queue files. Therefore, you must provide at least two import information queue files; the maximum is eight files. Do not neglect to specify at least qufile001 and qufile002.
  You must specify the qufile001 to qufile008 operands consecutively in ascending order beginning with qufile001. If they are not consecutive or in ascending order, Datareplicator uses only those import information queue files that are specified consecutively in ascending order from the beginning. Each filename must be unique in the target system.
  
• queuesize=import-information-queue-file-size
  ~ <unsigned integer> ((2-1000000000)) (KB)
  Specify the file size of the import information queue files specified with the qufile001 to qufile008 operands. The file size specified in the queuesize operand applies to each of the files specified with the qufile001 to qufile008 operands.
  For details about the formula for determining the size of an import information queue file, see 4.7.7 Designing the target Datareplicator's resources.
  If you handle the import information queue files as large files, specify at lease 2097152 (2 GB). To handle large files, you must already have set the OS and Datareplicator file system areas to support large files. For details, see 6.11 Handling of large files.
  

  Note:

  If you change the size of an existing import information queue file, you must initialize the target Datareplicator. Therefore, before you change file sizes, make sure that replication has been completed.

• reffile="import-definition-filename"
  ~ <[pathname/]filename of 1-64 bytes>
  Specify the name of the import definition file containing the definitions of import processing, as an absolute or relative pathname. If you specify a relative pathname, Datareplicator assumes $HDSPATH/relative-pathname as the absolute pathname. In this case, specify the pathname so that the final absolute pathname will not exceed 125 characters.
  If the source and target tables have exactly the same format, name, and column names, you can omit the reffile operand. The specified filename must be unique in the target system.
  
• statsfile="import-status-filename"
  ~ <[pathname/]filename of 1-64 bytes>
  Specify the name of the status file to be used to store information needed for error recovery processing, as an absolute or relative pathname. If you specify a relative pathname, Datareplicator assumes $HDSPATH/relative-pathname as the absolute pathname. In this case, specify the pathname so that the final absolute pathname will not exceed 125 characters. The specified filename must be unique in the target system.
  If you rename a status file but a file with the same name already exists, use the hdsstart -i -f command to execute an initial start of the target Datareplicator.
  
• statssize=import-status-file-size
  ~ <unsigned integer> ((72-2000000)) (KB)
  Specify the size of the import status file. For details about the size of the import status file, see 4.7.7 Designing the target Datareplicator's resources.
  
• unreffile1="unimported-information-filename-(primary)"unreffile2="unimported-information-filename-(secondary)"
  ~ <[pathname/]filename of 1-64 bytes>
  Specify the names of the unimported information files, as absolute or relative pathnames. If you specify a relative pathname, Datareplicator assumes $HDSPATH/relative-pathname as the absolute pathname. In this case, specify the pathname so that the final absolute pathname will not exceed 125 characters.
  There is a primary (unreffile1) and a secondary (unreffile2) unimported information file. Datareplicator uses these two files to accumulate information by swapping them. The specified filenames must be unique in the target system.
  
• unreffilesz=unimported-information-file-size
  ~ <unsigned integer> ((1-32767)) <<16>> (KB)
  Specify the maximum size of an unimported information file. This size applies to both unreffile1 and unreffile2. For details about the size of the unimported information files, see 4.7.7 Designing the target Datareplicator's resources.
  
• startmode={ trn|tbl|spd }
  Specify the import method to be used when the target Datareplicator starts. The import method specified in the startmode operand is used only when the previous termination mode was normal termination, event termination, or immediate termination. If the target Datareplicator was terminated forcibly by the hdsstop -t force command or because of an import processing error, the import method used before the termination is inherited.
  

  trn

  Start import processing using the transaction-based import method.

  tbl

  Start import processing using a table-based import method (table-based partitioning method, key range-based partitioning method, or hash partitioning method). If tbl is selected, a transaction occurs for each table's import group. Therefore, if the source system updates multiple tables with a single transaction, the target system updates the tables with separate transactions, which means that synchronization is not assured. If the tables must be synchronized, define them in the same import group.

  spd

  Start reception processing only and keep import processing inactive.

• restartmode=initial|continue
  Specify whether the import processing status at the time of the previous termination is to be inherited when the target Datareplicator starts (whether import processing is to be active or inactive).
  If the import processing was terminated by the hdsrfctl -m immediate command or due to detection of an import stop event, the previous import processing status is inherited. If the import processing was terminated by the hdsstop command or because of an import processing error, the previous import processing status is not inherited.
  

  initial

  Start import processing on the basis of the startmode operand's specification without inheriting the import processing status in effect at the time of the previous termination.

  continue

  Start import processing by inheriting the import processing status in effect at the time of the previous termination.

• breaktime=hh:mm
  Specify the amount of time (hours and minutes) that is to elapse from startup of the target Datareplicator to startup of import processing (applicable when spd is specified in the startmode operand). The amount of time specified is relative to startup of the target Datareplicator. Specifying more than 23:59 or less than 00:00 will result in an operand specification error. If you omit the breaktime operand or specify breaktime=00:00, import processing will not start until it is started by entry of the hdsrfctl command.
  

  hh ~ <unsigned integer> ((00-23))

  mm ~ <unsigned integer> ((00-59))

  hh (hours) and mm (minutes) must each be specified as two digits. For example, to start import processing after one hour and one minute, specify 01:01.

• breakmode=trn|tbl
  Specify the import method to be used when import processing is started after startup of the target Datareplicator. This operand is applicable when the following two conditions are satisfied and you have specified the breakmode operand:
  
  • spd is specified in the startmode operand
  • A valid breaktime operand specification is in effect

  trn

  Start import processing using the transaction-based import method.

  tbl

  Start import processing using a table-based import method (table-based partitioning method, key range-based partitioning method, or hash partitioning method).

• eventtrn=transaction-based-import-event-code
  ~ <unsigned integer> ((1-128)) <<1>>
  Specify the event code to be used to start import processing with the transaction-based import method. You must establish the correspondence of the transaction-based import event code in the source system.
  
• eventtbl=table-based-import-event-code
  ~ <unsigned integer> ((1-128)) <<2>>
  Specify the event code to be used to start import processing with a table-based import method (table-based partitioning method, key range-based partitioning method, or hash partitioning method). You must establish the correspondence of the table-based import event code in the source system.
  
• eventretrn=transaction-based-import-restart-event-code
  ~ <unsigned integer> ((1-128)) <<3>>
  Specify the event code to be used to restart import processing, which is inactive, with the transaction-based import method. You need to establish the correspondence of the transaction-based import restart event codes in the source system.
  
• eventretbl=table-based-import-restart-event-code
  ~ <unsigned integer> ((1-128)) <<4>>
  Specify the event code to be used to restart import processing, which is inactive, with a table-based import method (table-based partitioning method, key range-based partitioning method, or hash partitioning method). You must establish the correspondence of the table-based import restart event code in the source system.
  
• eventspd=import-processing-stop-event-code
  ~ <unsigned integer> ((1-128)) <<5>>
  Specify the event code to be used to stop import processing. You must establish the correspondence of the import processing stop event code in the source system.
  
• eventcntreset=import-processing-count-reset-event-code
  ~ <unsigned integer> ((1-128))
  Specify the event code used for resetting of the import processing count. If resetting is to be synchronized between the source and target Datareplicators, this operand's value must be the same as the eventcntreset operand's value in the transmission environment definition.
  This operand's value must be different from the value of any of the following operands (including their default values) in the import environment definition:
  
  • eventtrn operand
  • eventtbl operand
  • eventspd operand
  • eventretrn operand
  • eventretbl operand
• defmerge=true|false
  Specify whether import processing is to be executed using the applicable source system's extraction definition as the import definition when no update information is defined in the import definition. The defmerge operand is applicable when an import definition is available.
  

  true

  Execute import processing on all tables that correspond to the source definition. If there is no import definition corresponding to an extraction definition, use the extraction definition as the import definition to continue import processing.

  false

  Execute import processing only on the tables that are defined in the import definition.

• cmtintvl=import-processing-commit-interval
  ~ <unsigned integer> ((1-32767)) <<100>>
  Specify the interval at which the target Datareplicator is to issue commit to the target HiRDB, expressed as a number of transactions at the source system. In the following cases, the target Datareplicator always issues a commit to the target HiRDB regardless of the ccmtintvl operand specification:
  
  • An event is detected
  • PURGE TABLE is executed during import processing
  If you specify large values in the cmtintvl, trncmtintvl, and tblcmtintvl operands, the HiRDB server is adversely affected. For details, see 4.7.3(5) Designing the COMMIT-issuance interval for import processing.
  Also, when a large value is specified in the cmtintvl operand, a large amount of the following HiRDB resources is used because the number of SQL statements executed by one transaction on the target Datareplicator increases:
  
  • pd_lck_pool_size
  • pd_max_access_tables
  • pd_log_sdinterval
  • pd_sql_object_cache_size
  If an error occurs due to a shortage of locked resources, specify a smaller value in the cmtintvl operand.
  
• trncmtintvl=import-processing-commit-interval-for-transaction-based-import-method
  ~ <unsigned integer> ((1-32767)) <<100>>
  Specify the interval at which the target Datareplicator is to issue commit to the target HiRDB during import processing using the transaction-based import method, expressed as a number of transactions at the source system. In the following cases, the target Datareplicator always issues a commit to the target HiRDB regardless of the trncmtintvl operand specification:
  
  • An event is detected
  • PURGE TABLE is executed during import processing
  When the transaction-based import method is used, the trncmtintvl operand takes precedence over the cmtintvl operand.
  If you specify large values in the cmtintvl, trncmtintvl, and tblcmtintvl operands, the HiRDB server is adversely affected. For details, see 4.7.3(5) Designing the COMMIT-issuance interval for import processing.
  Also, if a large value is specified in the trncmtintvl operand, a large amount of the following HiRDB resources is used because the number of SQL statements executed by one transaction on the target Datareplicator increases:
  
  • pd_lck_pool_size
  • pd_max_access_tables
  • pd_log_sdinterval
  • pd_sql_object_cache_size
  If an error occurs due to a shortage of locked resources, specify a smaller value in the trncmtintvl operand.
  
• tblcmtintvl=import-processing-commit-interval-for-table-based-import-method
  ~ <unsigned integer> ((1-32767)) <<100>>
  Specify the interval at which the target Datareplicator is to issue commit to the target HiRDB during import processing using the table-based import method, expressed as a number of transactions at the source system. In the following cases, the target Datareplicator always issues a commit to the target HiRDB regardless of the tblcmtintvl operand specification:
  
  • An event is detected
  • PURGE TABLE is executed during import processing
  When the table-based import method is used, the tblcmtintvl operand takes precedence over the cmtintvl operand.
  If you specify large values in the cmtintvl, trncmtintvl, and tblcmtintvl operands, the HiRDB server is adversely affected. For details, see 4.7.3(5) Designing the COMMIT-issuance interval for import processing.
  Also, if a large value is specified in the tblcmtintvl operand, a large amount of the following HiRDB resources is used because the number of SQL statements executed by one transaction on the target Datareplicator increases:
  
  • pd_lck_pool_size
  • pd_max_access_tables
  • pd_log_sdinterval
  • pd_sql_object_cache_size
  If an error occurs due to a shortage of locked resources, specify a smaller value in the tblcmtintvl operand.
  
• tblcheck=true|false
  Specify whether the target system is to be checked when the target Datareplicator starts for the table subject to import processing. If an import definition is available, the target Datareplicator always checks to see if the table specified in the import definition exists at the target system, regardless of the tblcheck operand specification. If no import definitions are provided, the target Datareplicator checks to see if the table specified in the extraction definition sent from the source system exists at the target system only if true is specified in the tblcheck operand.
  

  true

  Check to see if the table subject to import processing exists at the target system. This might result in an increase in the workload at the target HiRDB when the target Datareplicator starts.

  false

  Do not check.

• defshmsize=shared-memory-size-for-storing-definition-information
  ~ <unsigned integer> ((1-2000000)) (KB)
  

  If you are using UNIX Datareplicator:

  Specify the size of the shared memory to be used to store the definition information analysis results. For details about the size of the shared memory, see 4.7.7 Designing the target Datareplicator's resources.

  If you are using Windows Datareplicator:

  Specify the size of the memory-mapped file to be used to store the definition information analysis results. For details about the size of the memory-mapped file, see 4.7.7 Designing the target Datareplicator's resources.

• ebcdic_type={ eck78|ekk78|eck83|ekk83 }
  Specify the EBCDIK/KEIS type for the data to be sent when the source system uses the EBCDIK/KEIS character code system or the source Datareplicator is to convert data to EBCDIK/KEIS. The ebcdic_type operand is applicable only when the transmitted data uses the EBCDIK/KEIS character code system.
  

  eck78

  Use EBCDIC/KEIS78.

  ekk78

  Use EBCDIK/KEIS78.

  eck83

  Use EBCDIC/KEIS83.

  ekk83

  Use EBCDIK/KEIS83.

• shiftspace_cnv=multi|single
  Specify the type of space character (double-byte or single-byte) to be used to replace two consecutive single-byte space characters in the double-byte mode. The shiftspace_cnv operand is applicable only when the transmitted data uses the EBCDIK/KEIS character code system.
  

  multi

  Convert two consecutive single-byte space characters to a single double-byte space character.

  single

  Convert two consecutive single-byte space characters to two consecutive single-byte space characters.

• undefcode_cnv=multi|single
  Specify the type of space character (double-byte or single-byte) to be used to replace an undefined Gaiji character.
  

  multi

  Convert an undefined Gaiji character to one double-byte space character.

  single

  Convert an undefined Gaiji character to two consecutive single-byte space characters.

• ref_data_backspace=suppress|nosuppress
  Specify suppress when all the conditions listed below are satisfied. If the defined lengths are the same at both the source and the target systems when suppress is not specified, an error results when an SQL statement is issued.
  The following table shows the conditions under which an error results when an SQL statement is issued, depending on the character code system used in the source and target databases:
  

  Extraction locale	Import locale
  	EBCDIK(C)/KEIS	SJIS	EUC	UTF-8
  EBCDIK(C)/KEIS	--	N	E A1 and B2 A2 and B1	E A1 and B3 A2 and B3 A3 and B3
  SJIS	--	--
  EUC	--	N	--
  UTF-8	--	N	N	--

  Legend:

  N: No error results when an SQL statement is issued.

  E: An error results when an SQL statement is issued:

  A1: Data in the target database contains single-byte Kana characters.

  A2: Data in the target database contains Gaiji characters.

  A3: Data in the target database contains non-ASCII code.

  B1: Column definition length in the target database does not exceed 1.5 times (3/2) the column definition length in the source database.

  B2: Column definition length in the target database does not exceed 2 times (2/1) the column definition length in the source database.

  B3: Column definition length in the target database does not exceed 3 times (3/1) the column definition length in the source database.

When suppress is specified, the CPU time might increase because Datareplicator deletes space characters after character code conversion. Therefore, if the source database's character code system is not EUC or UTF-8, we recommend that you specify nosuppress. If data consists of only space characters, Datareplicator regards the data as a single space character and executes import processing.

suppress

Eliminate all trailing space characters.

nosuppress

Do not eliminate any trailing space characters.

• skip_sqlcode=SQLCODE[,SQLCODE ] ...
  Specify the SQLCODEs for database updating errors (SQL errors) that are to be skipped (ignored) during import processing so that import processing will continue.
  The following two types of SQL errors can be skipped:
  
  • SQL errors that occur during update processing (EXECUTE)
  • SQL errors that occur during INSERT or UPDATE preprocessing (PREPARE)
  Note that Datareplicator versions earlier than 05-00 cannot skip SQL errors that occur during SQL preprocessing (PREPARE).
  You can specify a maximum of 32 SQLCODEs.
  To specify a negative SQLCODE, use the minus sign. If a specified SQLCODE is for an SQL error that is subject to implicit rollback, Datareplicator ignores the skip specification and stops import processing.
  This operand has no default value. If you omit this operand, Datareplicator will stop import processing whenever any SQL error occurs, regardless of the SQLCODE. For details about SQLCODEs, see the HiRDB Version 9 UAP Development Guide.
  
• skip_mvcelmwarn=true|false
  Specify the action to be taken by Datareplicator when UPDATE is executed on a table containing a repetition column during import processing and W is set in SQLWARN7 (the element number specified in the SET or DELETE clause of the UPDATE statement is not found in the row subject to update processing).
  

  true

  Skip SQLWARN7 and resume processing.

  false

  Handle it as an error and stop import processing.

• sqlerr_skip_info={ output|msgoutput|sqloutput|nooutput }
  Specify whether a warning message indicating a skip or an error is to be output and whether the unimported information associated with an SQL error is to be output according to the specification of the following operands:
  
  • skip_sqlcode operand specifying that import of update information associated with specified SQLCODEs is to be skipped
  • skip_mvcelmwarn operand specifying true (to skip warning and resume processing when W is set in SQLWARN7)
  This operand is applicable only when the SQLCODEs are specified in the skip_sqlcode operand or when true is specified in the skip_mvcelmwarn operand.
  

  output

  Output both a warning message and the unimported information.

  msgoutput

  Output a warning message only.

  sqloutput

  Output the unimported information only.

  nooutput

  Output neither a warning message nor the unimported information.

• extract_init=check|nocheck
  Specify whether the reception sequence of update information is to be checked when connection with the source system is established immediately after initialization.
  

  check

  Check that the source system is also initialized when connection is established with the source system immediately after initialization.

  nocheck

  Establish connection with the source system unconditionally after initialization.

• db_connect_retry_number=DB-connect-retries-count
  ~ <0-255> <<0>> (count)
  Specify the number of times connection establishment might be reattempted when the source Datareplicator's attempt to establish connection with the target database results in an error.
  
• db_connect_retry_interval=DB-connect-retry-interval
  ~ <0-3600> <<30>> (seconds)
  Specify in seconds the retry interval when the source Datareplicator attempts to establish connection with the target database. This operand is applicable only when 1 or a greater value is specified in the db_connect_retry_number operand.
  
• skip_codecnv_error=true|false
  Specify whether to stop import processing or to skip the update information containing erroneous extraction data and resume import processing when character code conversion on extraction data being processed for import results in an error.
  

  true

  In the event of an error, skip the update information containing the erroneous extraction data and resume import processing.

  false

  In the event of an error, stop import processing.

• int_trc_getl=activity-trace-collection-range
  Specify the range of activity trace information that is to be collected for each data linkage identifier at this system. This operand is applicable only when nc is specified in the second parameter of the int_trc_lvl operand in the import system definition.
  
  • Value for the activity trace collection range
    

    Value	Information to be collected
    	MST (control)	RCV (reception)	RFC (import)	SQE (SQL execution)
    c1	Y	Y	Y	--
    c2	Y	Y	--	Y
    c3	Y	Y	--	--
    c4	Y	--	Y	Y
    c5	Y	--	Y	--
    c6	Y	--	-	Y
    c7	Y	--	-	--
    activity-trace-collection-range omitted	Y	--	--	--

    Y: Collected.
    --: Not collected.
    
• device01=Datareplicator-file-system-area-name[[,allocation-file-type] ...]
  ~ <absolute pathname of 1-125 bytes>
  device02=Datareplicator-file-system-area-name[[,allocation-file-type] ...]
  ~ <absolute pathname of 1-125 bytes>
  :
  device09=Datareplicator-file-system-area-name[[,allocation-file-type] ...]
  ~ <absolute pathname of 1-125 bytes>
  Specify the names of the Datareplicator file system areas and the file types to be allocated in each file system area. Specify a two-digit sequence number for xx in devicexx. The following table shows the correspondence between files and the file types to which they are allocated:
  

  Allocation file type	File name specified in each operand
  qufilexxx (xxx: integer from 001 to 008)	Link file name for import information queue file specified in set qufilexxx
  statsfile	Link file name for import status file specified in set statsfile

  If you do not specify any allocation file types, all files at the corresponding dsid become subject to allocation. If a file system area is specified in a devicexx operand, Datareplicator automatically creates it and links it to the Datareplicator file system. You must ensure that the name of the file corresponding to an allocation file type has not already been allocated as a character special file.
  The following is a definition example of the devicexx operand:
  

  

  

Note:

To omit a devicexx operand from the definition after Datareplicator has been operated using the devicexx operand:

1. Stop Datareplicator.
2. Delete the devicexx operand from the import environment definition.
3. Use the hdsfmkfs command to initialize the Datareplicator file system area that was specified with the deleted devicexx operand.
4. Execute initial start on Datareplicator.

If you delete a devicexx operand from the definition, and then execute initial start on Datareplicator without following the above procedure, an invalid file error might occur during execution.

• ujcodekind=rcv|sam
  Specify whether the data linkage identifier is to be used with the normal reception facility or to input update information. When the target Datareplicator establishes data linkage with a mainframe database that uses SAM files (PDM2 E2 or RDB1 E2), you must specify sam in the ujcodekind operand.
  

  rcv

  Use the data linkage identifier with Datareplicator's normal reception facility.

  sam

  Use the data linkage identifier to input update information.

• discintvl=disconnect-issuance-interval
  ~ <unsigned integer> ((0-65535)) <<180>> (seconds)
  For details about this operand, see the discintvl operand in the import system definition. When this operand is specified, it takes precedence over the discintvl operand in the import system definition. If you omit this operand, the discintvl operand in the import system definition is effective. If you also omit the discintvl operand in the import system definition, Datareplicator assumes the default value for the discintvl operand in the import system definition.
  
• ref_wait_interval=import-process's-import-information-queue-file-read-interval
  ~ <unsigned integer> ((100-60000)) <<5000>> (milliseconds)
  For details about this operand, see the ref_wait_interval operand in the import system definition. When this operand is specified, it takes precedence over the ref_wait_interval operand in the import system definition. If you omit this operand, the ref_wait_interval operand in the import system definition is effective. If you also omit the ref_wait_interval operand in the import system definition, Datareplicator assumes the default value for the ref_wait_interval operand in the import system definition.
  
• commit_wait_time=COMMIT-issuance-interval
  ~ <unsigned integer> ((0-300)) <<30>> (seconds)
  For details about this operand, see the commit_wait_time operand in the import system definition. When this operand is specified, it takes precedence over the commit_wait_time operand in the import system definition. If you omit this operand, the commit_wait_time operand in the import system definition is effective. If you also omit the commit_wait_time operand in the import system definition, Datareplicator assumes the default value for the commit_wait_time operand in the import system definition.
  
• mapping_key_check= { not_null_unique|unique|none }
  Specify the unique condition for performing unique check on the mapping key column. This value becomes the default when the check clause is omitted in the import definition. Specification of this clause is effective when the check clause is specified in the import definition. For details about unique checking, see check clause in 5.10.5 Import table definition.
  

  not_null_unique

  Check the target table to confirm that the index satisfying the conditions described in Table 5-18 Details of uniqueness checking has been defined and its index component columns have the NOT NULL attribute.

  unique

  Check the target table to confirm that the index satisfying the conditions described in Table 5-18 Details of uniqueness checking has been defined. Because this option does not check the NULL value, you need to use a tool such as a user program that updates the source database to ensure the uniqueness of the data.

  none

  Do not perform a check. You need to use a tool such as a user program that updates the source database to ensure the uniqueness of the data.

  Note:

  If Datareplicator is not to perform checking to determine whether the target table is located at the target system, Datareplicator will not check the mapping key regardless of this operand's value. The checking to determine whether the target system contains the target table depends on the tblcheck operand in the import environment definition.

• control_trigger= execute|not_execute
  Specify whether a trigger set by the user is to execute on a target table.
  If a trigger is executed in the following cases, inconsistency might occur between the source and target databases:
  
  • A BEFORE trigger uses an assignment statement of a routine control SQL.
  • An AFTER trigger is defined for a target table that is to be replicated

  execute

  Execute the trigger on the target table.

  not_execute

  Do not execute a trigger on the target table.

  For details about data linkage for a table for which a trigger has been set, see 3.3.6 Data linkage for a table with a trigger set.
  
• control_reference_trigger= execute | not_execute
  Specify whether a trigger set by HiRDB for a target table is to be executed.
  

  execute

  Execute a trigger set for a target table.

  not_execute

  Do not execute a trigger set for a target table.

  If there is a referencing table with CASCADE specified as the referential constraint action, specify not_execute.

  For details about data linkage for a table for which a trigger has been set, see 3.3.6 Data linkage for a table with a trigger set.
  
• check_pending= use | nouse
  Specify whether a target table with a referential constraint defined is to be placed in check pending status. This operand takes effect only when not_execute is specified in the control_reference_trigger operand.
  

  use

  Place such a target table in check pending status.

  nouse

  Do not place such a target table in check pending status.

• reflect_delay_limit_time=import-delay-period-threshold
  ~ <unsigned integer> ((0-86400)) <<0>> (seconds)
  Specify the threshold value for the difference (import delay period) between the time at which update information is stored in the system log file and the time at which update information is imported into the target database. If the actual import delay period exceeds the threshold value specified here, a warning message is output.
  You must also treat the following operand values in the import environment definition as delay periods:
  
  • cmtintvl operand
  • trncmtintvl operand
  • tblcmtintvl operand
  • ref_wait_interval operand
  • commit_wait_time operand
  If a value of 0 is specified, the delay monitoring facility is disabled for import processes.
  

  Note:

  Adjust the value of this operand taking into account the amount of update information subject to replication, the transmission interval, and the import interval. A recommended guideline to start with is 20 minutes (1200 seconds).

  Example

  The following figure shows an example of operation when the start of transmission processing is delayed.

  Figure 5-8 Example of operation where the start of transmission processing is delayed

  
  

  This example executes only extraction processing from 8:00 to 17:00 and starts transmission processing after 17:00. To monitor the import delay period in such an operating environment, update data is not transmitted for up to 9 hours and is accumulated during that period at the target system. Therefore, a period of 9 hours must be added to the threshold for the import delay period.

• sql_lockerr_retrynum=transaction-retry-count-in-the-event-of-a-lock-error
  ~ <unsigned integer> ((0-255)) <<3>>
  Specify the retry count for executing a transaction that has rolled back for one of the reasons listed below. If a value of 0 is specified, there will be no retries.
  
  • Import processing resulted in a lock error.
  • A deadlock occurred.
  • A communication error occurred.
  In Datareplicator, there is no retry interval because a HiRDB lock-release wait time is specified in the pd_lck_wait_timeout operand in the HiRDB system command definition. This operand specifies the number of retries per target transaction; therefore, if the target transmission issues a commit, this retry count is reset.
  The following describes an example in which 10 is specified in this operand:
  
  • There will be 4 retries after a lock error occurs 4 times on a transaction.
  • The next transaction does not retry for 6 times. The retry count is reset for the next transaction. Therefore, each transaction has a maximum of 10 retries for successful execution.
  • The first lock error is not included in this retry count. If 10 is specified, the retry count is not exceeded until 11 lock errors occur.
  The following describes when retry occurs for each target DBMS:
  

  DBMS subject to import processing	When retry occurs	Target of retry
  HiRDB	In the event of an error, SQLCODE is as follows: -770 (lock-release timeout error) -911 (deadlock error) -722 (communication error) -723 (communication error)	Transaction that contains the update processing resulting in the error
  Oracle (Windows) SQL Server	ODBC function's return value is 40001 (deadlock error)	Transaction that contains the update processing resulting in the error
  Oracle (other than Windows)	OCI function's return value is 60 (deadlock error)	Update processing resulting in the error

• xa_recovery_retry_count=transaction-recovery-request-retries-count
  ~ <unsigned integer>((0-1024))<<180>>
  Specify the number of times a transaction recovery request is to be issued in the event transaction recovery fails when a transaction recovery request has been issued to HiRDB but HiRDB has not started or there is such a large load on HiRDB's acceptance processing that HiRDB might not have been able to recover the transaction.
  This operand takes effect when fxa_sqle is specified in the commitment_method operand in the import system definition.
  Specify an appropriate value using as a guideline the number of SQL processes in Datareplicator's import group (maximum number of SQL processes that can issue a recovery request concurrently to HiRDB). If 0 is specified, a transaction recovery request will not be reissued.
  
• xa_recovery_retry_interval=transaction-recovery-request-retry-interval
  ~ <unsigned integer>((1-180000))<<1000>>(milliseconds)
  Specify the interval (milliseconds) at which transaction recovery requests are to be issued in the event transaction recovery fails when a transaction recovery request has been issued to HiRDB but HiRDB has not started or there is such a large load on HiRDB's acceptance processing that HiRDB might not have been able to recover the transaction.
  This operand takes effect when fxa_sqle is specified in the commitment_method operand in the import system definition.
  Specifying a small value in this operand reduces the time required for import processing, but CPU usage by SQL processes increases and the overall system load might become large. Specify an appropriate value taking CPU performance into account.
  If the event of a large number of SQL processes in Datareplicator's import group (maximum number of SQL processes that can issue recovery requests concurrently to HiRDB), specify a larger retry interval to minimize the load created by HiRDB recovery processing.
  
• reflect_trn_max_sqlnum=maximum-update-SQL-statements-count-in-import-transaction
  ~ <unsigned integer> ((1-10000000))
  To import multiple source transactions as a single import transaction, specify the maximum number of update SQL statements in the import transaction. This value determines the number of source transactions to be treated as a single import transaction. If this operand is omitted, this value depends on the value of the cmtintvl, trncmtintvl, or tblcmtintvl operand.
  Also, if a large value is specified in the reflect_trn_max_sqlnum operand, a large amount of the following HiRDB resources is used because the number of SQL statements executed by one transaction on the target Datareplicator increases:
  
  • pd_lck_pool_size
  • pd_max_access_tables
  • pd_log_sdinterval
  • pd_sql_object_cache_size
  If an error occurs due to a shortage of locked resources, specify a smaller value in the reflect_trn_max_sqlnum operand.
  The number of source transactions to be treated as a single import transaction is determined as follows:
  

  Number of update SQL statements handled by the import transaction	Details of import transaction
  reflect_trn_max_sqlnum operand value or less	Determined by the cmtintvl, trncmtintvl, or tblcmtintvl operand
  Greater than the reflect_trn_max_sqlnum operand value	Until the total number of update SQL statements exceeds the specified value, all source transactions are handled as a single import transaction. If the specified number of update SQL statements is exceeded by a single source transaction, that transaction is handled as a single import transaction.

  The following figure shows an example in which the specifications are cmtintvl=2 and reflect_trn_max_sqlnum=20.
  

  Figure 5-9 Relationship between source and target transactions

  

  1. The number of update SQL statements for T1 is no greater than the reflect_trn_max_sqlnum operand value, and the cmtintvl operand value is 2. Therefore, whether T2 and T1 can be combined in order to be handled as a single target transaction is evaluated.
     The total number of update SQL statements for T1 and T2 is no greater than the reflect_trn_max_sqlnum operand value. Therefore, T1 and T2 can be combined in order to be handled as a single target transaction.
     
  2. The number of update SQL statements for T3 is no greater than the reflect_trn_max_sqlnum operand value, and the cmtintvl operand value is 2. Therefore, whether T3 and T4 can be combined in order to be handled as a single target transaction is evaluated.
     The total number of update SQL statements for T3 and T4 exceeds the reflect_trn_max_sqlnum operand value. Therefore, T3 and T4 are handled as separate target transactions.
     
  3. The remaining T5 is handled as a single target transaction.

  Note:

  As a guideline for the reflect_trn_max_sqlnum operand's value, specify the maximum number of resources required by the target DBMS to process a single transaction.