Exporting data to the file system

The tools dump command dumps data and information about data schema objects to the client file system in the format described in the File structure article:

ydb [connection options] tools dump [options]

where [connection options] are database connection options

[options]: Command parameters:

-p PATH or --path PATH: Path to the DB directory whose objects are to be dumped or path to the table. By default, the DB root directory. The following will be dumped: all subdirectories whose names do not begin with a dot and tables whose names do not begin with a dot inside these subdirectories. To dump such tables or the contents of such directories, you can explicitly specify their names in this parameter.

-o PATH or --output PATH: Path to the directory in the client file system that data should be dumped to. If the specified directory doesn't exist, it will be created. Anyway, the entire path to it must exist. If the specified directory does exist, it must be empty. If the parameter is not specified, a directory with a name in backup_YYYYDDMMTHHMMSS format will be created in the current directory, where YYYYDDMM indicates the date and HHMMSS the export start time.

--exclude STRING: Pattern (PCRE) for excluding paths from the export destination. This parameter can be specified several times for different patterns.

--scheme-only: Only dump information about data schema objects and no data.

--consistency-level VAL: Consistency level. Possible options:

  • database: Fully consistent export with a single snapshot taken before starting the export operation. Applied by default.
  • table: Consistency within each table being dumped with separate independent snapshots taken for each such table. It can run faster and have less impact on handling the current DB load.

--avoid-copy: Do not create a dump snapshot. The snapshot used by default to ensure consistency may not be applicable in some cases (such as for tables with external blobs).

--save-partial-result: Do not delete the result of a partially completed dump. If this option is not enabled, the result will be deleted in case an error occurs when dumping data.


The examples use a profile named db1. For information about how to create it, see the Getting started with the YDB CLI article in the "Getting started " section.

Exporting a database

With a directory named backup_... automatically created in the current directory:

ydb --profile db1 tools dump 

To the specified directory:

ydb --profile db1 tools dump -o ~/backup_db1

Exporting the structure of tables in the specified DB directory and its subdirectories

ydb --profile db1 tools dump -p dir1 --scheme-only