HiRDB Datareplicator Version 8 Description, User's Guide and Operator's Guide

[Contents][Glossary][Index][Back][Next]

3.7.5 Synchronous import group

To use the import transaction synchronization facility, you must set the target Datareplicator's data linkage identifiers as a synchronous import group in order to use a single target Datareplicator to receive the transaction that is transmitted by each back-end server of the source Datareplicator.

Organization of this subsection
(1) Starting a synchronous import group
(2) Terminating a synchronous import group
(3) Notes about synchronous import groups

(1) Starting a synchronous import group

To start a synchronous import group, specify the required operands correctly, and then enter one of the following commands:

  1. hdsstart
  2. hdsrfctl -g synchronous-import-group-name -m start

The command shown in 2 can be executed only when the synchronization managing process is inactive. If you execute this command while the synchronization managing process is running, the KFRB00161-E message is output and the command is ignored. You can use the hdsstate command to check whether the process is running.

To start a synchronous import group in disabled status, enter the following command:

hdsstart -c synchronous-import-group-name

Once you enter this command, the next time the command shown in 1 or 2 is entered, the synchronous import group is started in disabled status. To enable the synchronous import group again, you must initialize the target Datareplicator.

Note
A synchronous import group cannot be started for an individual data linkage identifier that constitutes the group. If the hdsrfctl -d command is executed on a data linkage identifier constituting a synchronous import group, the KFRB03304-E message is output and the command is ignored.

(2) Terminating a synchronous import group

To terminate a synchronous import group, enter one of the following commands:

#
For details about the available options, see the hdsstop command in Chapter 7. Command Syntax.

The following table describes the termination method for each command.

Table 3-20 How a synchronous import group is terminated

Command specification Termination details
hdsstop The reception process has stopped and the end of the import information queue file has been detected. If a transaction was executing when the end of the import information queue file was detected, this command rolls back that transaction, and then terminates the synchronous import group.
hdsstop -t event
hdsstop -q wait-time
hdsstop -t immediate Completion of synchronization at the first synchronous event after acceptance of the command, or when the end of the import information queue file is detected. If a transaction was executed when the end of the import information queue file was detected, this command rolls back that transaction, and then terminates the synchronous import group.
hdsrfctl -g synchronous-import-group-name -m immediate
hdsstop -t force Rolls back the import transaction that was executing when this command was accepted, and then terminates the synchronous import group. If no import transaction has occurred, the command terminates the synchronous import group as is.
Note
A synchronous import group cannot be terminated for an individual data linkage identifier that constitutes the group. If the hdsrfctl -d command is executed on a data linkage identifier constituting a synchronous import group, the KFRB03304-E message is output and the command is ignored.

(3) Notes about synchronous import groups

The notes in this subsection apply to synchronous import groups.

(a) Notes about operand specification

There are limitations on the operand specifications for the data linkage identifiers specified for a synchronous import group, as described in the following table.

Table 3-21 Limitations on the data linkage identifiers specified for a synchronous import group

Definition file name Operand Remarks
Import system definition commit_wait_time Ignored
commitment_method Specify fxa_sqle. Otherwise, the KFRB00847-E message is output, resulting in an error.
discintvl Ignored
Import environment definition eventtrn Ignored
eventtbl Ignored
eventretrn Ignored
eventretbl Ignored
eventspd Ignored
eventcntreset Ignored
cmtintvl Ignored
trncmtintvl Ignored
tblcmtintvl Ignored
reflect_trn_max_sqlnum Ignored
startmode# Specify trn. Otherwise, the KFRB00847-E message is output, resulting in an error.
ujcodekind Specify rcv. Otherwise, the KFRB00847-E message is output, resulting in an error.
Import definition load statement by 'uoc-name' cannot be specified. If it is specified, the KFRB03315-E message is output, resulting in an error.

#
Use the startmode operand to specify the method for importing the data linkage identifiers specified in a synchronous import group. The import method cannot be changed by the event issuing facility or the hdsrfctl command.

You must evaluate the specification values beforehand, because starting a synchronous import group in disabled status enables all the parameters, including those that had been ignored.

(b) Notes about the event issuing facility

Among the events for a synchronous import group, only synchronous events are available. An attempt to use any other type of event results in output of the KFRB03316-W message, and the attempt is ignored.

A connection-based termination event is also ignored; it will not be treated as a -1 event.

(c) Notes about errors in a synchronous import group

If there is an error in any of the data linkage identifiers that constitute a synchronous import group, all the data linkage identifiers constituting that synchronous import group terminate with an error.

When an error occurs in a synchronous import group, the synchronization managing process is terminated and each import process that detected termination of the synchronization managing process outputs the KFRB03311-E message.

Termination of a synchronization managing process is detected at one of the following times:

(d) Notes about the number of transactions that is displayed by the hdsstate command

The number of transactions displayed for each data linkage identifier by the hdsstate command is the number of transaction branches. It does not match the number of actual global transactions that have occurred at the source system.