33.1.1 Functional overview

The adbmergechunk command is used to merge multiple chunks in multi-chunk tables into one new chunk. This is called merge chunk processing.

You can reduce the number of chunks in a table by performing merge chunk processing.

The following figure provides an overview of merge chunk processing.

Figure 33‒1: Overview of merge chunk processing

Explanation:: The adbmergechunk command merges table data that exists in any number of chunks (merge-source chunks) into a single new chunk (merge-target chunk). The command then rebuilds the indexes defined for the table data in the merge-target chunk. Finally, the command deletes the merge-source chunks. This allows you to use a single chunk to manage table data and indexes that had previously been managed in multiple chunks.

Important: You cannot merge chunks that are in a different status (chunks in normal status and chunks in wait status).

Note: As the number of chunks increases, the performance of retrievals using B-tree or text indexes might decrease. You can avoid such deterioration of retrieval performance by merging chunks and reducing their number.