25.5.23 ndb_restore — Restore an NDB Cluster Backup

The NDB Cluster restoration program is implemented as a separate command-line utility ndb_restore, which can normally be found in the MySQL bin directory. This program reads the files created as a result of the backup and inserts the stored information into the database.

ndb_restore must be executed once for each of the backup files that were created by the START BACKUP command used to create the backup (see Section 25.6.8.2, “Using The NDB Cluster Management Client to Create a Backup”). This is equal to the number of data nodes in the cluster at the time that the backup was created.

Note

Before using ndb_restore, it is recommended that the cluster be running in single user mode, unless you are restoring multiple data nodes in parallel. See Section 25.6.6, “NDB Cluster Single User Mode”, for more information.

Options that can be used with ndb_restore are shown in the following table. Additional descriptions follow the table.

• --allow-pk-changes

  Command-Line Format	--allow-pk-changes[=0|1]
  Type	Integer
  Default Value	0
  Minimum Value	0
  Maximum Value	1

  When this option is set to 1, ndb_restore allows the primary keys in a table definition to differ from that of the same table in the backup. This may be desirable when backing up and restoring between different schema versions with primary key changes on one or more tables, and it appears that performing the restore operation using ndb_restore is simpler or more efficient than issuing many ALTER TABLE statements after restoring table schemas and data.

  The following changes in primary key definitions are supported by --allow-pk-changes:

  • Extending the primary key: A non-nullable column that exists in the table schema in the backup becomes part of the table's primary key in the database.

    Important

    When extending a table's primary key, any columns which become part of primary key must not be updated while the backup is being taken; any such updates discovered by ndb_restore cause the restore operation to fail, even when no change in value takes place. In some cases, it may be possible to override this behavior using the --ignore-extended-pk-updates option; see the description of this option for more information.

  • Contracting the primary key (1): A column that is already part of the table's primary key in the backup schema is no longer part of the primary key, but remains in the table.

  • Contracting the primary key (2): A column that is already part of the table's primary key in the backup schema is removed from the table entirely.

  These differences can be combined with other schema differences supported by ndb_restore, including changes to blob and text columns requiring the use of staging tables.

  Basic steps in a typical scenario using primary key schema changes are listed here:

  1. Restore table schemas using ndb_restore --restore-meta

  2. Alter schema to that desired, or create it

  3. Back up the desired schema

  4. Run ndb_restore --disable-indexes using the backup from the previous step, to drop indexes and constraints

  5. Run ndb_restore --allow-pk-changes (possibly along with --ignore-extended-pk-updates, --disable-indexes, and possibly other options as needed) to restore all data

  6. Run ndb_restore --rebuild-indexes using the backup made with the desired schema, to rebuild indexes and constraints

  When extending the primary key, it may be necessary for ndb_restore to use a temporary secondary unique index during the restore operation to map from the old primary key to the new one. Such an index is created only when necessary to apply events from the backup log to a table which has an extended primary key. This index is named NDB$RESTORE_PK_MAPPING, and is created on each table requiring it; it can be shared, if necessary, by multiple instances of ndb_restore instances running in parallel. (Running ndb_restore --rebuild-indexes at the end of the restore process causes this index to be dropped.)

• --append

  Command-Line Format	--append

  When used with the --tab and --print-data options, this causes the data to be appended to any existing files having the same names.

• --backup-path=dir_name

  Command-Line Format	--backup-path=path
  Type	Directory name
  Default Value	./

  The path to the backup directory is required; this is supplied to ndb_restore using the --backup-path option, and must include the subdirectory corresponding to the ID backup of the backup to be restored. For example, if the data node's DataDir is /var/lib/mysql-cluster, then the backup directory is /var/lib/mysql-cluster/BACKUP, and the backup files for the backup with the ID 3 can be found in /var/lib/mysql-cluster/BACKUP/BACKUP-3. The path may be absolute or relative to the directory in which the ndb_restore executable is located, and may be optionally prefixed with backup-path=.

  It is possible to restore a backup to a database with a different configuration than it was created from. For example, suppose that a backup with backup ID 12, created in a cluster with two storage nodes having the node IDs 2 and 3, is to be restored to a cluster with four nodes. Then ndb_restore must be run twice—once for each storage node in the cluster where the backup was taken. However, ndb_restore cannot always restore backups made from a cluster running one version of MySQL to a cluster running a different MySQL version. See Section 25.3.7, “Upgrading and Downgrading NDB Cluster”, for more information.

  Important

  It is not possible to restore a backup made from a newer version of NDB Cluster using an older version of ndb_restore. You can restore a backup made from a newer version of MySQL to an older cluster, but you must use a copy of ndb_restore from the newer NDB Cluster version to do so.

  For example, to restore a cluster backup taken from a cluster running NDB Cluster 8.4.0 to a cluster running NDB Cluster 8.0.41, you must use the ndb_restore that comes with the NDB Cluster 8.0.41 distribution.

  For more rapid restoration, the data may be restored in parallel, provided that there is a sufficient number of cluster connections available. That is, when restoring to multiple nodes in parallel, you must have an [api] or [mysqld] section in the cluster config.ini file available for each concurrent ndb_restore process. However, the data files must always be applied before the logs.

• --backup-password=password

  Command-Line Format	--backup-password=password
  Type	String
  Default Value	[none]

  This option specifies a password to be used when decrypting an encrypted backup with the --decrypt option. This must be the same password that was used to encrypt the backup.

  The password must be 1 to 256 characters in length, and must be enclosed by single or double quotation marks. It can contain any of the ASCII characters having character codes 32, 35, 38, 40-91, 93, 95, and 97-126; in other words, it can use any printable ASCII characters except for !, ', ", $, %, \, and ^.

  It is possible to omit the password, in which case ndb_restore waits for it to be supplied from stdin, as when using --backup-password-from-stdin.

• --backup-password-from-stdin[=TRUE|FALSE]

  Command-Line Format	--backup-password-from-stdin

  When used in place of --backup-password, this option enables input of the backup password from the system shell (stdin), similar to how this is done when supplying the password interactively to mysql when using the --password without supplying the password on the command line.

• --backupid=#, -b

  Command-Line Format	--backupid=#
  Type	Numeric
  Default Value	none

  This option is required; it is used to specify the ID or sequence number of the backup, and is the same number shown by the management client in the Backup backup_id completed message displayed upon completion of a backup. (See Section 25.6.8.2, “Using The NDB Cluster Management Client to Create a Backup”.)

  Important

  When restoring cluster backups, you must be sure to restore all data nodes from backups having the same backup ID. Using files from different backups results at best in restoring the cluster to an inconsistent state, and is likely to fail altogether.

• --character-sets-dir

  Command-Line Format	--character-sets-dir=path

  Directory containing character sets.

• --connect, -c

  Command-Line Format	--connect=connection_string
  Type	String
  Default Value	localhost:1186

  Alias for --ndb-connectstring.

• --connect-retries

  Command-Line Format	--connect-retries=#
  Type	Integer
  Default Value	12
  Minimum Value	0
  Maximum Value	12

  Number of times to retry connection before giving up.

• --connect-retry-delay

  Command-Line Format	--connect-retry-delay=#
  Type	Integer
  Default Value	5
  Minimum Value	0
  Maximum Value	5

  Number of seconds to wait between attempts to contact management server.

• --connect-string

  Command-Line Format	--connect-string=connection_string
  Type	String
  Default Value	[none]

  Same as --ndb-connectstring.

• --core-file

  Command-Line Format	--core-file

  Write core file on error; used in debugging.

• --decrypt

  Command-Line Format	--decrypt

  Decrypt an encrypted backup using the password supplied by the --backup-password option.

• --defaults-extra-file

  Command-Line Format	--defaults-extra-file=path
  Type	String
  Default Value	[none]

  Read given file after global files are read.

• --defaults-file

  Command-Line Format	--defaults-file=path
  Type	String
  Default Value	[none]

  Read default options from given file only.

• --defaults-group-suffix

  Command-Line Format	--defaults-group-suffix=string
  Type	String
  Default Value	[none]

  Also read groups with concat(group, suffix).

• --disable-indexes

  Command-Line Format	--disable-indexes

  Disable restoration of indexes during restoration of the data from a native NDB backup. Afterwards, you can restore indexes for all tables at once with multithreaded building of indexes using --rebuild-indexes, which should be faster than rebuilding indexes concurrently for very large tables.

  This option also drops any foreign keys specified in the backup.

  MySQL can open an NDB table for which one or more indexes cannot be found, provided the query does not use any of the affected indexes; otherwise the query is rejected with ER_NOT_KEYFILE. In the latter case, you can temporarily work around the problem by executing an ALTER TABLE statement such as this one:

  ALTER TABLE tbl ALTER INDEX idx INVISIBLE;

  This causes MySQL to ignore the index idx on table tbl. See Primary Keys and Indexes, for more information, as well as Section 10.3.12, “Invisible Indexes”.

• --dont-ignore-systab-0, -f

  Command-Line Format	--dont-ignore-systab-0

  Normally, when restoring table data and metadata, ndb_restore ignores the copy of the NDB system table that is present in the backup. --dont-ignore-systab-0 causes the system table to be restored. This option is intended for experimental and development use only, and is not recommended in a production environment.

• --exclude-databases=db-list

  Command-Line Format	--exclude-databases=list
  Type	String
  Default Value

  Comma-delimited list of one or more databases which should not be restored.

  This option is often used in combination with --exclude-tables; see that option's description for further information and examples.

• --exclude-intermediate-sql-tables[=TRUE|FALSE]

  Command-Line Format	--exclude-intermediate-sql-tables[=TRUE|FALSE]
  Type	Boolean
  Default Value	TRUE

  When performing copying ALTER TABLE operations, mysqld creates intermediate tables (whose names are prefixed with #sql-). When TRUE, the --exclude-intermediate-sql-tables option keeps ndb_restore from restoring such tables that may have been left over from these operations. This option is TRUE by default.

• --exclude-missing-columns

  Command-Line Format	--exclude-missing-columns

  It is possible to restore only selected table columns using this option, which causes ndb_restore to ignore any columns missing from tables being restored as compared to the versions of those tables found in the backup. This option applies to all tables being restored. If you wish to apply this option only to selected tables or databases, you can use it in combination with one or more of the --include-* or --exclude-* options described elsewhere in this section to do so, then restore data to the remaining tables using a complementary set of these options.

• --exclude-missing-tables

  Command-Line Format	--exclude-missing-tables

  It is possible to restore only selected tables using this option, which causes ndb_restore to ignore any tables from the backup that are not found in the target database.

• --exclude-tables=table-list

  Command-Line Format	--exclude-tables=list
  Type	String
  Default Value

  List of one or more tables to exclude; each table reference must include the database name. Often used together with --exclude-databases.

  When --exclude-databases or --exclude-tables is used, only those databases or tables named by the option are excluded; all other databases and tables are restored by ndb_restore.

  This table shows several invocations of ndb_restore using --exclude-* options (other options possibly required have been omitted for clarity), and the effects these options have on restoring from an NDB Cluster backup:

  Table 25.23 Several invocations of ndb_restore using --exclude-* options, and the effects these options have on restoring from an NDB Cluster backup.

  Option	Result
  --exclude-databases=db1	All tables in all databases except db1 are restored; no tables in db1 are restored
  --exclude-databases=db1,db2 (or --exclude-databases=db1 --exclude-databases=db2)	All tables in all databases except db1 and db2 are restored; no tables in db1 or db2 are restored
  --exclude-tables=db1.t1	All tables except t1 in database db1 are restored; all other tables in db1 are restored; all tables in all other databases are restored
  --exclude-tables=db1.t2,db2.t1 (or --exclude-tables=db1.t2 --exclude-tables=db2.t1)	All tables in database db1 except for t2 and all tables in database db2 except for table t1 are restored; no other tables in db1 or db2 are restored; all tables in all other databases are restored

  
  

  You can use these two options together. For example, the following causes all tables in all databases except for databases db1 and db2, and tables t1 and t2 in database db3, to be restored:

  $> ndb_restore [...] --exclude-databases=db1,db2 --exclude-tables=db3.t1,db3.t2

  (Again, we have omitted other possibly necessary options in the interest of clarity and brevity from the example just shown.)

  You can use --include-* and --exclude-* options together, subject to the following rules:

  • The actions of all --include-* and --exclude-* options are cumulative.

  • All --include-* and --exclude-* options are evaluated in the order passed to ndb_restore, from right to left.

  • In the event of conflicting options, the first (rightmost) option takes precedence. In other words, the first option (going from right to left) that matches against a given database or table “wins”.

  For example, the following set of options causes ndb_restore to restore all tables from database db1 except db1.t1, while restoring no other tables from any other databases:

  --include-databases=db1 --exclude-tables=db1.t1

  However, reversing the order of the options just given simply causes all tables from database db1 to be restored (including db1.t1, but no tables from any other database), because the --include-databases option, being farthest to the right, is the first match against database db1 and thus takes precedence over any other option that matches db1 or any tables in db1:

  --exclude-tables=db1.t1 --include-databases=db1

• --fields-enclosed-by=char

  Command-Line Format	--fields-enclosed-by=char
  Type	String
  Default Value

  Each column value is enclosed by the string passed to this option (regardless of data type; see the description of --fields-optionally-enclosed-by).

• --fields-optionally-enclosed-by

  Command-Line Format	--fields-optionally-enclosed-by
  Type	String
  Default Value

  The string passed to this option is used to enclose column values containing character data (such as CHAR, VARCHAR, BINARY, TEXT, or ENUM).

• --fields-terminated-by=char

  Command-Line Format	--fields-terminated-by=char
  Type	String
  Default Value	\t (tab)

  The string passed to this option is used to separate column values. The default value is a tab character (\t).

• --help

  Command-Line Format	--help

  Display help text and exit.

• --hex

  Command-Line Format	--hex

  If this option is used, all binary values are output in hexadecimal format.

• --ignore-extended-pk-updates

  Command-Line Format	--ignore-extended-pk-updates[=0|1]
  Type	Integer
  Default Value	0
  Minimum Value	0
  Maximum Value	1

  When using --allow-pk-changes, columns which become part of a table's primary key must not be updated while the backup is being taken; such columns should keep the same values from the time values are inserted into them until the rows containing the values are deleted. If ndb_restore encounters updates to these columns when restoring a backup, the restore fails. Because some applications may set values for all columns when updating a row, even when some column values are not changed, the backup may include log events appearing to update columns which are not in fact modified. In such cases you can set --ignore-extended-pk-updates to 1, forcing ndb_restore to ignore such updates.

  Important

  When causing these updates to be ignored, the user is responsible for ensuring that there are no updates to the values of any columns that become part of the primary key.

  For more information, see the description of --allow-pk-changes.

• --include-databases=db-list

  Command-Line Format	--include-databases=list
  Type	String
  Default Value

  Comma-delimited list of one or more databases to restore. Often used together with --include-tables; see the description of that option for further information and examples.

• --include-stored-grants

  Command-Line Format	--include-stored-grants

  ndb_restore does not by default restore shared users and grants (see Section 25.6.13, “Privilege Synchronization and NDB_STORED_USER”) to the ndb_sql_metadata table. Specifying this option causes it to do so.

• --include-tables=table-list

  Command-Line Format	--include-tables=list
  Type	String
  Default Value

  Comma-delimited list of tables to restore; each table reference must include the database name.

  When --include-databases or --include-tables is used, only those databases or tables named by the option are restored; all other databases and tables are excluded by ndb_restore, and are not restored.

  The following table shows several invocations of ndb_restore using --include-* options (other options possibly required have been omitted for clarity), and the effects these have on restoring from an NDB Cluster backup:

  Table 25.24 Several invocations of ndb_restore using --include-* options, and their effects on restoring from an NDB Cluster backup.

  Option	Result
  --include-databases=db1	Only tables in database db1 are restored; all tables in all other databases are ignored
  --include-databases=db1,db2 (or --include-databases=db1 --include-databases=db2)	Only tables in databases db1 and db2 are restored; all tables in all other databases are ignored
  --include-tables=db1.t1	Only table t1 in database db1 is restored; no other tables in db1 or in any other database are restored
  --include-tables=db1.t2,db2.t1 (or --include-tables=db1.t2 --include-tables=db2.t1)	Only the table t2 in database db1 and the table t1 in database db2 are restored; no other tables in db1, db2, or any other database are restored

  
  

  You can also use these two options together. For example, the following causes all tables in databases db1 and db2, together with the tables t1 and t2 in database db3, to be restored (and no other databases or tables):

  $> ndb_restore [...] --include-databases=db1,db2 --include-tables=db3.t1,db3.t2

  (Again we have omitted other, possibly required, options in the example just shown.)

  It also possible to restore only selected databases, or selected tables from a single database, without any --include-* (or --exclude-*) options, using the syntax shown here:

  ndb_restore other_options db_name,[db_name[,...] | tbl_name[,tbl_name][,...]]

  In other words, you can specify either of the following to be restored:

  • All tables from one or more databases

  • One or more tables from a single database

• --lines-terminated-by=char

  Command-Line Format	--lines-terminated-by=char
  Type	String
  Default Value	\n (linebreak)

  Specifies the string used to end each line of output. The default is a linefeed character (\n).

• --login-path

  Command-Line Format	--login-path=path
  Type	String
  Default Value	[none]

  Read given path from login file.

• --no-login-paths

  Command-Line Format	--no-login-paths

  Skips reading options from the login path file.

• --lossy-conversions, -L

  Command-Line Format	--lossy-conversions

  This option is intended to complement the --promote-attributes option. Using --lossy-conversions allows lossy conversions of column values (type demotions or changes in sign) when restoring data from backup. With some exceptions, the rules governing demotion are the same as for MySQL replication; see Section 19.5.1.9.2, “Replication of Columns Having Different Data Types”, for information about specific type conversions currently supported by attribute demotion.

  This option also makes it possible to restore a NULL column as NOT NULL. The column must not contain any NULL entries; otherwise ndb_restore stops with an error.

  ndb_restore reports any truncation of data that it performs during lossy conversions once per attribute and column.

• --no-binlog

  Command-Line Format	--no-binlog

  This option prevents any connected SQL nodes from writing data restored by ndb_restore to their binary logs.

• --no-restore-disk-objects, -d

  Command-Line Format	--no-restore-disk-objects

  This option stops ndb_restore from restoring any NDB Cluster Disk Data objects, such as tablespaces and log file groups; see Section 25.6.11, “NDB Cluster Disk Data Tables”, for more information about these.

• --no-upgrade, -u

  Command-Line Format	--no-upgrade

  When using ndb_restore to restore a backup, VARCHAR columns created using the old fixed format are resized and recreated using the variable-width format now employed. This behavior can be overridden by specifying --no-upgrade.

• --ndb-connectstring

  Command-Line Format	--ndb-connectstring=connection_string
  Type	String
  Default Value	[none]

  Set connection string for connecting to ndb_mgmd. Syntax: [nodeid=id;][host=]hostname[:port]. Overrides entries in NDB_CONNECTSTRING and my.cnf.

• --ndb-mgm-tls

  Command-Line Format	--ndb-mgm-tls=level
  Type	Enumeration
  Default Value	relaxed
  Valid Values	relaxed strict

  Sets the level of TLS support required to connect to the management server; one of relaxed or strict. relaxed (the default) means that a TLS connection is attempted, but success is not required; strict means that TLS is required to connect.

• --ndb-mgmd-host

  Command-Line Format	--ndb-mgmd-host=connection_string
  Type	String
  Default Value	[none]

  Same as --ndb-connectstring.

• --ndb-nodegroup-map=map, -z

  Command-Line Format	--ndb-nodegroup-map=map

  Any value set for this option is ignored, and the option itself does nothing.

• --ndb-nodeid

  Command-Line Format	--ndb-nodeid=#
  Type	Integer
  Default Value	[none]

  Set node ID for this node, overriding any ID set by --ndb-connectstring.

• --ndb-optimized-node-selection

  Command-Line Format	--ndb-optimized-node-selection

  Enable optimizations for selection of nodes for transactions. Enabled by default; use --skip-ndb-optimized-node-selection to disable.

• --ndb-tls-search-path

  Command-Line Format	--ndb-tls-search-path=list
  Type	Path name
  Default Value (Unix)	$HOME/ndb-tls
  Default Value (Windows)	$HOMEDIR/ndb-tls

  Specify a list of directories to search for a CA file. On Unix platforms, the directory names are separated by colons (:); on Windows systems, the semicolon character (;) is used as the separator. A directory reference may be relative or absolute; it may contain one or more environment variables, each denoted by a prefixed dollar sign ($), and expanded prior to use.

  Searching begins with the leftmost named directory and proceeds from left to right until a file is found. An empty string denotes an empty search path, which causes all searches to fail. A string consisting of a single dot (.) indicates that the search path limited to the current working directory.

  If no search path is supplied, the compiled-in default value is used. This value depends on the platform used: On Windows, this is \ndb-tls; on other platforms (including Linux), it is $HOME/ndb-tls. This can be overridden by compiling NDB Cluster using -DWITH_NDB_TLS_SEARCH_PATH.

• --no-defaults

  Command-Line Format	--no-defaults

  Do not read default options from any option file other than login file.

• --nodeid=#, -n

  Command-Line Format	--nodeid=#
  Type	Numeric
  Default Value	none

  Specify the node ID of the data node on which the backup was taken; required.

  When restoring to a cluster with different number of data nodes from that where the backup was taken, this information helps identify the correct set or sets of files to be restored to a given node. (In such cases, multiple files usually need to be restored to a single data node.) See Restoring to a different number of data nodes, for additional information and examples.

• --num-slices=#

  Command-Line Format	--num-slices=#
  Type	Integer
  Default Value	1
  Minimum Value	1
  Maximum Value	1024

  When restoring a backup by slices, this option sets the number of slices into which to divide the backup. This allows multiple instances of ndb_restore to restore disjoint subsets in parallel, potentially reducing the amount of time required to perform the restore operation.

  A slice is a subset of the data in a given backup; that is, it is a set of fragments having the same slice ID, specified using the --slice-id option. The two options must always be used together, and the value set by --slice-id must always be less than the number of slices.

  ndb_restore encounters fragments and assigns each one a fragment counter. When restoring by slices, a slice ID is assigned to each fragment; this slice ID is in the range 0 to 1 less than the number of slices. For a table that is not a BLOB table, the slice to which a given fragment belongs is determined using the formula shown here:

  [slice_ID] = [fragment_counter] % [number_of_slices]

  For a BLOB table, a fragment counter is not used; the fragment number is used instead, along with the ID of the main table for the BLOB table (recall that NDB stores BLOB values in a separate table internally). In this case, the slice ID for a given fragment is calculated as shown here:

  [slice_ID] =
  ([main_table_ID] + [fragment_ID]) % [number_of_slices]

  Thus, restoring by N slices means running N instances of ndb_restore, all with --num-slices=N (along with any other necessary options) and one each with --slice-id=1, --slice-id=2, --slice-id=3, and so on through slice-id=N-1.

  Example.  Assume that you want to restore a backup named BACKUP-1, found in the default directory /var/lib/mysql-cluster/BACKUP/BACKUP-3 on the node file system on each data node, to a cluster with four data nodes having the node IDs 1, 2, 3, and 4. To perform this operation using five slices, execute the sets of commands shown in the following list:

  1. Restore the cluster metadata using ndb_restore as shown here:

     $> ndb_restore -b 1 -n 1 -m --disable-indexes --backup-path=/home/ndbuser/backups

  2. Restore the cluster data to the data nodes invoking ndb_restore as shown here:

     $> ndb_restore -b 1 -n 1 -r --num-slices=5 --slice-id=0 --backup-path=/var/lib/mysql-cluster/BACKUP/BACKUP-1
     $> ndb_restore -b 1 -n 1 -r --num-slices=5 --slice-id=1 --backup-path=/var/lib/mysql-cluster/BACKUP/BACKUP-1
     $> ndb_restore -b 1 -n 1 -r --num-slices=5 --slice-id=2 --backup-path=/var/lib/mysql-cluster/BACKUP/BACKUP-1
     $> ndb_restore -b 1 -n 1 -r --num-slices=5 --slice-id=3 --backup-path=/var/lib/mysql-cluster/BACKUP/BACKUP-1
     $> ndb_restore -b 1 -n 1 -r --num-slices=5 --slice-id=4 --backup-path=/var/lib/mysql-cluster/BACKUP/BACKUP-1
     
     $> ndb_restore -b 1 -n 2 -r --num-slices=5 --slice-id=0 --backup-path=/var/lib/mysql-cluster/BACKUP/BACKUP-1
     $> ndb_restore -b 1 -n 2 -r --num-slices=5 --slice-id=1 --backup-path=/var/lib/mysql-cluster/BACKUP/BACKUP-1
     $> ndb_restore -b 1 -n 2 -r --num-slices=5 --slice-id=2 --backup-path=/var/lib/mysql-cluster/BACKUP/BACKUP-1
     $> ndb_restore -b 1 -n 2 -r --num-slices=5 --slice-id=3 --backup-path=/var/lib/mysql-cluster/BACKUP/BACKUP-1
     $> ndb_restore -b 1 -n 2 -r --num-slices=5 --slice-id=4 --backup-path=/var/lib/mysql-cluster/BACKUP/BACKUP-1
     
     $> ndb_restore -b 1 -n 3 -r --num-slices=5 --slice-id=0 --backup-path=/var/lib/mysql-cluster/BACKUP/BACKUP-1
     $> ndb_restore -b 1 -n 3 -r --num-slices=5 --slice-id=1 --backup-path=/var/lib/mysql-cluster/BACKUP/BACKUP-1
     $> ndb_restore -b 1 -n 3 -r --num-slices=5 --slice-id=2 --backup-path=/var/lib/mysql-cluster/BACKUP/BACKUP-1
     $> ndb_restore -b 1 -n 3 -r --num-slices=5 --slice-id=3 --backup-path=/var/lib/mysql-cluster/BACKUP/BACKUP-1
     $> ndb_restore -b 1 -n 3 -r --num-slices=5 --slice-id=4 --backup-path=/var/lib/mysql-cluster/BACKUP/BACKUP-1
     
     $> ndb_restore -b 1 -n 4 -r --num-slices=5 --slice-id=0 --backup-path=/var/lib/mysql-cluster/BACKUP/BACKUP-1
     $> ndb_restore -b 1 -n 4 -r --num-slices=5 --slice-id=1 --backup-path=/var/lib/mysql-cluster/BACKUP/BACKUP-1
     $> ndb_restore -b 1 -n 4 -r --num-slices=5 --slice-id=2 --backup-path=/var/lib/mysql-cluster/BACKUP/BACKUP-1
     $> ndb_restore -b 1 -n 4 -r --num-slices=5 --slice-id=3 --backup-path=/var/lib/mysql-cluster/BACKUP/BACKUP-1
     $> ndb_restore -b 1 -n 4 -r --num-slices=5 --slice-id=4 --backup-path=/var/lib/mysql-cluster/BACKUP/BACKUP-1

     All of the commands just shown in this step can be executed in parallel, provided there are enough slots for connections to the cluster (see the description for the --backup-path option).

  3. Restore indexes as usual, as shown here:

     $> ndb_restore -b 1 -n 1 --rebuild-indexes --backup-path=/var/lib/mysql-cluster/BACKUP/BACKUP-1

  4. Finally, restore the epoch, using the command shown here:

     $> ndb_restore -b 1 -n 1 --restore-epoch --backup-path=/var/lib/mysql-cluster/BACKUP/BACKUP-1

  You should use slicing to restore the cluster data only; it is not necessary to employ --num-slices or --slice-id when restoring the metadata, indexes, or epoch information. If either or both of these options are used with the ndb_restore options controlling restoration of these, the program ignores them.

  The effects of using the --parallelism option on the speed of restoration are independent of those produced by slicing or parallel restoration using multiple instances of ndb_restore (--parallelism specifies the number of parallel transactions executed by a single ndb_restore thread), but it can be used together with either or both of these. You should be aware that increasing --parallelism causes ndb_restore to impose a greater load on the cluster; if the system can handle this, restoration should complete even more quickly.

  The value of --num-slices is not directly dependent on values relating to hardware such as number of CPUs or CPU cores, amount of RAM, and so forth, nor does it depend on the number of LDMs.

  It is possible to employ different values for this option on different data nodes as part of the same restoration; doing so should not in and of itself produce any ill effects.

• --parallelism=#, -p

  Command-Line Format	--parallelism=#
  Type	Numeric
  Default Value	128
  Minimum Value	1
  Maximum Value	1024

  ndb_restore uses single-row transactions to apply many rows concurrently. This parameter determines the number of parallel transactions (concurrent rows) that an instance of ndb_restore tries to use. By default, this is 128; the minimum is 1, and the maximum is 1024.

  The work of performing the inserts is parallelized across the threads in the data nodes involved. This mechanism is employed for restoring bulk data from the .Data file—that is, the fuzzy snapshot of the data; it is not used for building or rebuilding indexes. The change log is applied serially; index drops and builds are DDL operations and handled separately. There is no thread-level parallelism on the client side of the restore.

• --preserve-trailing-spaces, -P

  Command-Line Format	--preserve-trailing-spaces

  Cause trailing spaces to be preserved when promoting a fixed-width character data type to its variable-width equivalent—that is, when promoting a CHAR column value to VARCHAR, or a BINARY column value to VARBINARY. Otherwise, any trailing spaces are dropped from such column values when they are inserted into the new columns.

  Note

  Although you can promote CHAR columns to VARCHAR and BINARY columns to VARBINARY, you cannot promote VARCHAR columns to CHAR or VARBINARY columns to BINARY.

• --print

  Command-Line Format	--print

  Causes ndb_restore to print all data, metadata, and logs to stdout. Equivalent to using the --print-data, --print-meta, and --print-log options together.

  Note

  Use of --print or any of the --print_* options is in effect performing a dry run. Including one or more of these options causes any output to be redirected to stdout; in such cases, ndb_restore makes no attempt to restore data or metadata to an NDB Cluster.

• --print-data

  Command-Line Format	--print-data

  Cause ndb_restore to direct its output to stdout. Often used together with one or more of --tab, --fields-enclosed-by, --fields-optionally-enclosed-by, --fields-terminated-by, --hex, and --append.

  TEXT and BLOB column values are always truncated. Such values are truncated to the first 256 bytes in the output. This cannot currently be overridden when using --print-data.

• --print-defaults

  Command-Line Format	--print-defaults

  Print program argument list and exit.

• --print-log

  Command-Line Format	--print-log

  Cause ndb_restore to output its log to stdout.

• --print-meta

  Command-Line Format	--print-meta

  Print all metadata to stdout.

• print-sql-log

  Command-Line Format	--print-sql-log

  Log SQL statements to stdout. Use the option to enable; normally this behavior is disabled. The option checks before attempting to log whether all the tables being restored have explicitly defined primary keys; queries on a table having only the hidden primary key implemented by NDB cannot be converted to valid SQL.

  This option does not work with tables having BLOB columns.

• --progress-frequency=N

  Command-Line Format	--progress-frequency=#
  Type	Numeric
  Default Value	0
  Minimum Value	0
  Maximum Value	65535

  Print a status report each N seconds while the backup is in progress. 0 (the default) causes no status reports to be printed. The maximum is 65535.

• --promote-attributes, -A

  Command-Line Format	--promote-attributes

  ndb_restore supports limited attribute promotion in much the same way that it is supported by MySQL replication; that is, data backed up from a column of a given type can generally be restored to a column using a “larger, similar” type. For example, data from a CHAR(20) column can be restored to a column declared as VARCHAR(20), VARCHAR(30), or CHAR(30); data from a MEDIUMINT column can be restored to a column of type INT or BIGINT. See Section 19.5.1.9.2, “Replication of Columns Having Different Data Types”, for a table of type conversions currently supported by attribute promotion.

  This option also makes it possible to restore a NOT NULL column as NULL.

  Attribute promotion by ndb_restore must be enabled explicitly, as follows:

  1. Prepare the table to which the backup is to be restored. ndb_restore cannot be used to re-create the table with a different definition from the original; this means that you must either create the table manually, or alter the columns which you wish to promote using ALTER TABLE after restoring the table metadata but before restoring the data.

  2. Invoke ndb_restore with the --promote-attributes option (short form -A) when restoring the table data. Attribute promotion does not occur if this option is not used; instead, the restore operation fails with an error.

  When converting between character data types and TEXT or BLOB, only conversions between character types (CHAR and VARCHAR) and binary types (BINARY and VARBINARY) can be performed at the same time. For example, you cannot promote an INT column to BIGINT while promoting a VARCHAR column to TEXT in the same invocation of ndb_restore.

  Converting between TEXT columns using different character sets is not supported, and is expressly disallowed.

  When performing conversions of character or binary types to TEXT or BLOB with ndb_restore, you may notice that it creates and uses one or more staging tables named table_name$STnode_id. These tables are not needed afterwards, and are normally deleted by ndb_restore following a successful restoration.

• --rebuild-indexes

  Command-Line Format	--rebuild-indexes

  Enable multithreaded rebuilding of the ordered indexes while restoring a native NDB backup. The number of threads used for building ordered indexes by ndb_restore with this option is controlled by the BuildIndexThreads data node configuration parameter and the number of LDMs.

  It is necessary to use this option only for the first run of ndb_restore; this causes all ordered indexes to be rebuilt without using --rebuild-indexes again when restoring subsequent nodes. You should use this option prior to inserting new rows into the database; otherwise, it is possible for a row to be inserted that later causes a unique constraint violation when trying to rebuild the indexes.

  Building of ordered indices is parallelized with the number of LDMs by default. Offline index builds performed during node and system restarts can be made faster using the BuildIndexThreads data node configuration parameter; this parameter has no effect on dropping and rebuilding of indexes by ndb_restore, which is performed online.

  Rebuilding of unique indexes uses disk write bandwidth for redo logging and local checkpointing. An insufficient amount of this bandwidth can lead to redo buffer overload or log overload errors. In such cases you can run ndb_restore --rebuild-indexes again; the process resumes at the point where the error occurred. You can also do this when you have encountered temporary errors. You can repeat execution of ndb_restore --rebuild-indexes indefinitely; you may be able to stop such errors by reducing the value of --parallelism. If the problem is insufficient space, you can increase the size of the redo log (FragmentLogFileSize node configuration parameter), or you can increase the speed at which LCPs are performed (MaxDiskWriteSpeed and related parameters), in order to free space more quickly.

• --remap-column=db.tbl.col:fn:args

  Command-Line Format	--remap-column=string
  Type	String
  Default Value	[none]

  When used together with --restore-data, this option applies a function to the value of the indicated column. Values in the argument string are listed here:

  • db: Database name, following any renames performed by --rewrite-database.

  • tbl: Table name.

  • col: Name of the column to be updated. This column must be of type INT or BIGINT. The column can also be but is not required to be UNSIGNED.

  • fn: Function name; currently, the only supported name is offset.

  • args: Arguments supplied to the function. Currently, only a single argument, the size of the offset to be added by the offset function, is supported. Negative values are supported. The size of the argument cannot exceed that of the signed variant of the column's type; for example, if col is an INT column, then the allowed range of the argument passed to the offset function is -2147483648 to 2147483647 (see Section 13.1.2, “Integer Types (Exact Value) - INTEGER, INT, SMALLINT, TINYINT, MEDIUMINT, BIGINT”).

    If applying the offset value to the column would cause an overflow or underflow, the restore operation fails. This could happen, for example, if the column is a BIGINT, and the option attempts to apply an offset value of 8 on a row in which the column value is 4294967291, since 4294967291 + 8 = 4294967299 > 4294967295.

  This option can be useful when you wish to merge data stored in multiple source instances of NDB Cluster (all using the same schema) into a single destination NDB Cluster, using NDB native backup (see Section 25.6.8.2, “Using The NDB Cluster Management Client to Create a Backup”) and ndb_restore to merge the data, where primary and unique key values are overlapping between source clusters, and it is necessary as part of the process to remap these values to ranges that do not overlap. It may also be necessary to preserve other relationships between tables. To fulfill such requirements, it is possible to use the option multiple times in the same invocation of ndb_restore to remap columns of different tables, as shown here:

  $> ndb_restore --restore-data --remap-column=hr.employee.id:offset:1000 \
      --remap-column=hr.manager.id:offset:1000 --remap-column=hr.firstaiders.id:offset:1000

  (Other options not shown here may also be used.)

  --remap-column can also be used to update multiple columns of the same table. Combinations of multiple tables and columns are possible. Different offset values can also be used for different columns of the same table, like this:

  $> ndb_restore --restore-data --remap-column=hr.employee.salary:offset:10000 \
      --remap-column=hr.employee.hours:offset:-10

  When source backups contain duplicate tables which should not be merged, you can handle this by using --exclude-tables, --exclude-databases, or by some other means in your application.

  Information about the structure and other characteristics of tables to be merged can obtained using SHOW CREATE TABLE; the ndb_desc tool; and MAX(), MIN(), LAST_INSERT_ID(), and other MySQL functions.

  Replication of changes from merged to unmerged tables, or from unmerged to merged tables, in separate instances of NDB Cluster is not supported.

• --restore-data, -r

  Command-Line Format	--restore-data

  Output NDB table data and logs.

• --restore-epoch, -e

  Command-Line Format	--restore-epoch

  Add (or restore) epoch information to the cluster replication status table. This is useful for starting replication on an NDB Cluster replica. When this option is used, the row in the mysql.ndb_apply_status having 0 in the id column is updated if it already exists; such a row is inserted if it does not already exist. (See Section 25.7.9, “NDB Cluster Backups With NDB Cluster Replication”.)

• --restore-meta, -m

  Command-Line Format	--restore-meta

  This option causes ndb_restore to print NDB table metadata.

  The first time you run the ndb_restore restoration program, you also need to restore the metadata. In other words, you must re-create the database tables—this can be done by running it with the --restore-meta (-m) option. Restoring the metadata need be done only on a single data node; this is sufficient to restore it to the entire cluster.

  ndb_restore uses the default number of partitions for the target cluster, unless the number of local data manager threads is also changed from what it was for data nodes in the original cluster.

  When using this option, it is recommended that auto synchronization be disabled by setting ndb_metadata_check=OFF until ndb_restore has completed restoring the metadata, after which it can it turned on again to synchronize objects newly created in the NDB dictionary.

  Note

  The cluster should have an empty database when starting to restore a backup. (In other words, you should start the data nodes with --initial prior to performing the restore.)

• --restore-privilege-tables

  Command-Line Format	--restore-privilege-tables
  Deprecated	Yes

  No longer used.

• --rewrite-database=olddb,newdb

  Command-Line Format	--rewrite-database=string
  Type	String
  Default Value	none

  This option makes it possible to restore to a database having a different name from that used in the backup. For example, if a backup is made of a database named products, you can restore the data it contains to a database named inventory, use this option as shown here (omitting any other options that might be required):

  $> ndb_restore --rewrite-database=product,inventory

  The option can be employed multiple times in a single invocation of ndb_restore. Thus it is possible to restore simultaneously from a database named db1 to a database named db2 and from a database named db3 to one named db4 using --rewrite-database=db1,db2 --rewrite-database=db3,db4. Other ndb_restore options may be used between multiple occurrences of --rewrite-database.

  In the event of conflicts between multiple --rewrite-database options, the last --rewrite-database option used, reading from left to right, is the one that takes effect. For example, if --rewrite-database=db1,db2 --rewrite-database=db1,db3 is used, only --rewrite-database=db1,db3 is honored, and --rewrite-database=db1,db2 is ignored. It is also possible to restore from multiple databases to a single database, so that --rewrite-database=db1,db3 --rewrite-database=db2,db3 restores all tables and data from databases db1 and db2 into database db3.

  Important

  When restoring from multiple backup databases into a single target database using --rewrite-database, no check is made for collisions between table or other object names, and the order in which rows are restored is not guaranteed. This means that it is possible in such cases for rows to be overwritten and updates to be lost.

• --skip-broken-objects

  Command-Line Format	--skip-broken-objects

  This option causes ndb_restore to ignore corrupt tables while reading a native NDB backup, and to continue restoring any remaining tables (that are not also corrupted). Currently, the --skip-broken-objects option works only in the case of missing blob parts tables.

• --skip-table-check, -s

  Command-Line Format	--skip-table-check

  It is possible to restore data without restoring table metadata. By default when doing this, ndb_restore fails with an error if a mismatch is found between the table data and the table schema; this option overrides that behavior.

  Some of the restrictions on mismatches in column definitions when restoring data using ndb_restore are relaxed; when one of these types of mismatches is encountered, ndb_restore does not stop with an error as it did previously, but rather accepts the data and inserts it into the target table while issuing a warning to the user that this is being done. This behavior occurs whether or not either of the options --skip-table-check or --promote-attributes is in use. These differences in column definitions are of the following types:

  • Different COLUMN_FORMAT settings (FIXED, DYNAMIC, DEFAULT)

  • Different STORAGE settings (MEMORY, DISK)

  • Different default values

  • Different distribution key settings

• --skip-unknown-objects

  Command-Line Format	--skip-unknown-objects

  This option causes ndb_restore to ignore any schema objects it does not recognize while reading a native NDB backup. This can be used for restoring a backup made from a cluster running (for example) NDB 7.6 to a cluster running NDB Cluster 7.5.

• --slice-id=#

  Command-Line Format	--slice-id=#
  Type	Integer
  Default Value	0
  Minimum Value	0
  Maximum Value	1023

  When restoring by slices, this is the ID of the slice to restore. This option is always used together with --num-slices, and its value must be always less than that of --num-slices.

  For more information, see the description of the --num-slices elsewhere in this section.

• --tab=dir_name, -T dir_name

  Command-Line Format	--tab=path
  Type	Directory name

  Causes --print-data to create dump files, one per table, each named tbl_name.txt. It requires as its argument the path to the directory where the files should be saved; use . for the current directory.

• --timestamp-printouts

  Command-Line Format	--timestamp-printouts{=true|false}
  Type	Boolean
  Default Value	true

  Causes info, error, and debug log messages to be prefixed with timestamps.

  This option is enabled by default. Disable it with --timestamp-printouts=false.

• --usage

  Command-Line Format	--usage

  Display help text and exit; same as --help.

• --verbose=#

  Command-Line Format	--verbose=#
  Type	Numeric
  Default Value	1
  Minimum Value	0
  Maximum Value	255

  Sets the level for the verbosity of the output. The minimum is 0; the maximum is 255. The default value is 1.

• --version

  Command-Line Format	--version

  Display version information and exit.

• --with-apply-status

  Command-Line Format	--with-apply-status

  Restore all rows from the backup's ndb_apply_status table (except for the row having server_id = 0, which is generated using --restore-epoch). This option requires that --restore-data also be used.

  If the ndb_apply_status table from the backup already contains a row with server_id = 0, ndb_restore --with-apply-status deletes it. For this reason, we recommend that you use ndb_restore --restore-epoch after invoking ndb_restore with the --with-apply-status option. You can also use --restore-epoch concurrently with the last of any invocations of ndb_restore --with-apply-status used to restore the cluster.

  For more information, see ndb_apply_status Table.

Typical options for this utility are shown here:

ndb_restore [-c connection_string] -n node_id -b backup_id \
      [-m] -r --backup-path=/path/to/backup/files

Normally, when restoring from an NDB Cluster backup, ndb_restore requires at a minimum the --nodeid (short form: -n), --backupid (short form: -b), and --backup-path options.

The -c option is used to specify a connection string which tells ndb_restore where to locate the cluster management server (see Section 25.4.3.3, “NDB Cluster Connection Strings”). If this option is not used, then ndb_restore attempts to connect to a management server on localhost:1186. This utility acts as a cluster API node, and so requires a free connection “slot” to connect to the cluster management server. This means that there must be at least one [api] or [mysqld] section that can be used by it in the cluster config.ini file. It is a good idea to keep at least one empty [api] or [mysqld] section in config.ini that is not being used for a MySQL server or other application for this reason (see Section 25.4.3.7, “Defining SQL and Other API Nodes in an NDB Cluster”).

ndb_restore can decrypt an encrypted backup using --decrypt and --backup-password. Both options must be specified to perform decryption. See the documentation for the START BACKUP management client command for information on creating encrypted backups.

You can verify that ndb_restore is connected to the cluster by using the SHOW command in the ndb_mgm management client. You can also accomplish this from a system shell, as shown here:

$> ndb_mgm -e "SHOW"

Error reporting.  ndb_restore reports both temporary and permanent errors. In the case of temporary errors, it may able to recover from them, and reports Restore successful, but encountered temporary error, please look at configuration in such cases.

Important

After using ndb_restore to initialize an NDB Cluster for use in circular replication, binary logs on the SQL node acting as the replica are not automatically created, and you must cause them to be created manually. To cause the binary logs to be created, issue a SHOW TABLES statement on that SQL node before running START REPLICA. This is a known issue in NDB Cluster.