11.4.2 Storing data in a multi-chunk table (background import)

This subsection explains how to store data in a multi-chunk table by using background import.

Note

In addition to using background import, you can also use a method that stores data in a multi-chunk table in addition mode. For details about data import in addition mode, see 11.1.7 Storing data in a base table (data import).

To perform background import, execute the adbimport command with the -b option specified. The following shows an example of performing background import by specifying the -b option for the adbimport command.

For the specification format of the adbimport command, see adbimport (Import Data) in the manual HADB Command Reference.

■ Background import execution example

The HADB user (ADBUSER01) executes background import on the shop table (SHOPSLIST).

adbimport -u ADBUSER01 -p '#HelloHADB_01' -k "'" -s , -g 10
          -w /home/adbmanager/tmp
          -z /home/adbmanager/imp_file/imp_opt_file.txt
          -b -m '2014/01/01-2014/01/31'
          SHOPSLIST
          /home/adbmanager/imp_file/imp_data_path.txt

Note

■ Adding a comment to a chunk

When you import data into a multi-chunk table, you can add a comment to the chunk that will store the data. You use the -m option to add a comment.

When the operation targets a specific chunk, adding a comment to that chunk will make it easier to identify the target chunk.

Important

When you perform background import, for the adb_import_rthd_num import option, specify a value that satisfies the following formula. Specifying a value that satisfies the following formula for the adb_import_rthd_num import option will improve the data storage efficiency.

Formula

value-specified-for-import-option-adb_import_rthd_num ≤
      ↑ MIN ( C × db_org  , db_org × index_num × 0.25 ) ↑ + 1

Explanation of the variables

C

If the background import target table is a row store table, substitute 2. If the target table is a column store table, substitute 0.5.

db_org

File size of the input data file to be used for background import (gigabytes)

index_num

Number of B-tree indexes and text indexes that are defined for the background import target table

When background import executes, HADB creates a new chunk and stores the import data in that chunk. The following figure shows where data is stored by background import.

Figure 11‒3: Data storage destination by background import

Explanation

When the adbimport command is executed with the background-import facility, data is stored in newly created chunk 3. During this processing, the current chunk is changed from chunk 2 to chunk 3.

While data is being stored in chunk 3, you can retrieve the data stored in chunk 1 and chunk 2.

Important

Each time background import is executed, a new chunk is created. Consequently, repeating background import increases the number of chunks. When the number of chunks increases, performance might decline during retrieval processing using B-tree indexes and text indexes. Therefore, consider reducing the number of chunks by merging them. To merge chunks, see 11.4.9 Merging chunks (to reduce the number of chunks).

To Page Top