Preparing for data linkage recovery via unload log files

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

9.6.3 Preparing for data linkage recovery via unload log files

The preparations described in this subsection are required to perform data linkage recovery via unload log files. Make sure that all environment variables whose settings are required for the target Datareplicator have been set.

Organization of this subsection: (1) Creating the environment variable definition file (required for a node containing a back-end server subject to error recovery); (2) Preparing the unload log files; (3) Checking the available disk space (required for a node that contains a back-end server subject to error recovery); (4) Deleting or moving files created during the previous execution (applicable only if recovery processing is executed more than once on the same back-end server)

(1) Creating the environment variable definition file (required for a node containing a back-end server subject to error recovery)

Under the source Datareplicator directory ($HDEPATH), create the file named hde_toolenv that defines the following environment variable. Even if you omit the environment variable, create the hde_toolenv file as an empty file.

[TOOL_OUTPUT_DIR=execution-results-storage-directory-for-data-linkage-recovery-facility]

Note: Make sure that the environment variable's specification begins in column 1.

TOOL_OUTPUT_DIR specifies the directory used to store the files that are output by the data linkage recovery facility (such as a transactions list). If this environment variable is omitted, $HDEPATH is assumed. If a relative path is specified, Datareplicator assumes the path from $HDEPATH.

(2) Preparing the unload log files

At the applicable back-end server, prepare the HiRDB unload log files that are to be used for the recovery of data linkage. Also from the current system log file, obtain the unload log file after swapping, if necessary.

Store the prepared HiRDB unload log files in the same directory.

(3) Checking the available disk space (required for a node that contains a back-end server subject to error recovery)

Check the following items:

Make sure there is enough disk space for the unload log files at the target location. Also make sure there is enough disk space for the unload log files under the target directory that is specified when the logmrg command is executed.
Make sure there is enough disk space for storing the transactions list (under the source Datareplicator directory) that is created when data linkage recovery facility 1 is executed. The formula for determining the size of the transactions list is as follows:

Size of transactions list (bytes) = number of transactions to be recovered x 64 + 4

Obtain the number of transactions to be recovered from the sum of the results of executing the following command for all unload log files that are subject to recovery:

 
pdlogcat -s -kp unload-log-file-name|
     grep "Record header" | wc -l

Make sure that there is about 1 megabyte of free disk space under the $HDEPATH directory.

(4) Deleting or moving files created during the previous execution (applicable only if recovery processing is executed more than once on the same back-end server)

When data linkage recovery via unload log files is performed, the following files are created under the directory that is specified in the TOOL_OUTPUT_DIR operand in the environment variable definition file (hde_toolenv):

tranlist_BES-name: Target transactions list file
res_file_BES-name: Recovery-range unresolved transactions check file

If you execute data linkage recovery via unload log files more than once, check for the existence of these files. If they exist, either delete them or re-name them.