33.1.2 Relationships among threads used by the adbmergechunk command

The adbmergechunk command uses processing real threads in the following processing:

1. Retrieving and merging table data and creating range indexes

   One or more processing real threads are used.

2. Creating temporary work files

   One or more processing real threads are used.

3. Sorting records of B-tree and text indexes

   One or more processing real threads are used.

4. Creating B-tree and text indexes

   One or more processing real threads are used.

5. Deleting merge-source chunks

   One or more processing real threads are used.

6. Outputting the progress of command processing

   One processing real thread is used.

Steps 1 and 2 are performed at the same time. Then steps 3 to 5 are performed in sequence. Step 6 is always performed while steps 1 to 5 are in progress.

The following two figures show the processing of the adbmergechunk command using processing real threads. The preceding numbers correspond to the numbers in the figure.

Figure 33‒2: Processing of the adbmergechunk command using processing real threads (1)

Figure 33‒3: Processing of the adbmergechunk command using processing real threads (2)

The number of processing real threads used by the adbmergechunk command for each step is automatically calculated from the number of processing real threads to be used by the adbmergechunk command. For details about the number of processing real threads to be used by the adbmergechunk command, see the description of the merge chunk option adb_mergechunk_rthd_num in (2) Explanation of specification format in 33.2.2 Format of merge chunk options.

To Page Top