25.6.8.2 Using The NDB Cluster Management Client to Create a Backup (original) (raw)

25.6.8.2 Using The NDB Cluster Management Client to Create a Backup

Before starting a backup, make sure that the cluster is properly configured for performing one. (SeeSection 25.6.8.3, “Configuration for NDB Cluster Backups”.)

The START BACKUP command is used to create a backup, and has the syntax shown here:

START BACKUP [backup_id]
    [encryption_option]
    [wait_option]
    [snapshot_option]

encryption_option:
ENCRYPT [PASSWORD=password]

password:
{'password_string' | "password_string"}

wait_option:
WAIT {STARTED | COMPLETED} | NOWAIT

snapshot_option:
SNAPSHOTSTART | SNAPSHOTEND

Successive backups are automatically identified sequentially, so the backupid, an integer greater than or equal to 1, is optional; if it is omitted, the next available value is used. If an existing_backupid_ value is used, the backup fails with the error Backup failed: file already exists. If used, the_backupid_ must follow immediately after the START BACKUP keywords, before any other options are used.

In NDB 8.0.22 and later, START BACKUP supports the creation of encrypted backups usingENCRYPT PASSWORD=_`password`_. The_password_ must meet all of the following requirements:

When ENCRYPT PASSWORD='_`password`_' is used, the backup data record and log files written by each data node are encrypted with a key derived from the user-provided_password_ and a randomly-generated salt using a key derivation function (KDF) that employs the PBKDF2-SHA256 algorithm to generate a symmetric encryption key for that file. This function has the form shown here:

key = KDF(random_salt, password)

The key so generated is then used to encrypt the backup data using AES 256 CBC inline, and symmetric encryption is employed for encrypting the backup fileset (with the generated key).

Note

NDB Cluster never saves the user-furnished password or generated encryption key.

Starting with NDB 8.0.24, the PASSWORD option can be omitted from_encryptionoption_. In this case, the management client prompts the user for a password.

It is possible using PASSWORD to set an empty password ('' or ""), but this is not recommended.

An encrypted backup can be decrypted using any of the following commands:

NDB 8.0.24 and later supports the additional commands listed here:

See the descriptions of these programs for more information, such as additional options that may be required.

The waitoption can be used to determine when control is returned to the management client after a START BACKUP command is issued, as shown in the following list:

ndb_mgm> START BACKUP NOWAIT  
ndb_mgm>

In this case, the management client can be used even while it prints progress information from the backup process.

ndb_mgm> START BACKUP WAIT STARTED  
Waiting for started, this may take several minutes  
Node 2: Backup 3 started from node 1  
ndb_mgm>

WAIT COMPLETED is the default.

A snapshotoption can be used to determine whether the backup matches the state of the cluster when START BACKUP was issued, or when it was completed. SNAPSHOTSTART causes the backup to match the state of the cluster when the backup began;SNAPSHOTEND causes the backup to reflect the state of the cluster when the backup was finished.SNAPSHOTEND is the default, and matches the behavior found in previous NDB Cluster releases.

Note

If you use the SNAPSHOTSTART option withSTART BACKUP, and theCompressedBackup parameter is enabled, only the data and control files are compressed—the log file is not compressed.

If both a waitoption and a_snapshotoption_ are used, they may be specified in either order. For example, all of the following commands are valid, assuming that there is no existing backup having 4 as its ID:

START BACKUP WAIT STARTED SNAPSHOTSTART
START BACKUP SNAPSHOTSTART WAIT STARTED
START BACKUP 4 WAIT COMPLETED SNAPSHOTSTART
START BACKUP SNAPSHOTEND WAIT COMPLETED
START BACKUP 4 NOWAIT SNAPSHOTSTART

The procedure for creating a backup consists of the following steps:

  1. Start the management client (ndb_mgm), if it not running already.
  2. Execute the START BACKUP command. This produces several lines of output indicating the progress of the backup, as shown here:
ndb_mgm> START BACKUP  
Waiting for completed, this may take several minutes  
Node 2: Backup 1 started from node 1  
Node 2: Backup 1 started from node 1 completed  
 StartGCP: 177 StopGCP: 180  
 #Records: 7362 #LogRecords: 0  
 Data: 453648 bytes Log: 0 bytes  
ndb_mgm>
  1. When the backup has started the management client displays this message:
Backup backup_id started from node node_id

backupid is the unique identifier for this particular backup. This identifier is saved in the cluster log, if it has not been configured otherwise. nodeid is the identifier of the management server that is coordinating the backup with the data nodes. At this point in the backup process the cluster has received and processed the backup request. It does not mean that the backup has finished. An example of this statement is shown here: 

Node 2: Backup 1 started from node 1
  1. The management client indicates with a message like this one that the backup has started:
Backup backup_id started from node node_id completed

As is the case for the notification that the backup has started, backupid is the unique identifier for this particular backup, and_nodeid_ is the node ID of the management server that is coordinating the backup with the data nodes. This output is accompanied by additional information including relevant global checkpoints, the number of records backed up, and the size of the data, as shown here: 

Node 2: Backup 1 started from node 1 completed  
 StartGCP: 177 StopGCP: 180  
 #Records: 7362 #LogRecords: 0  
 Data: 453648 bytes Log: 0 bytes

It is also possible to perform a backup from the system shell by invoking ndb_mgm with the -e or --execute option, as shown in this example:

$> ndb_mgm -e "START BACKUP 6 WAIT COMPLETED SNAPSHOTSTART"

When using START BACKUP in this way, you must specify the backup ID.

Cluster backups are created by default in theBACKUP subdirectory of theDataDir on each data node. This can be overridden for one or more data nodes individually, or for all cluster data nodes in theconfig.ini file using theBackupDataDir configuration parameter. The backup files created for a backup with a given backupid are stored in a subdirectory namedBACKUP-_`backupid`_ in the backup directory.

Cancelling backups. To cancel or abort a backup that is already in progress, perform the following steps:

  1. Start the management client.
  2. Execute this command:
ndb_mgm> ABORT BACKUP backup_id

The number backupid is the identifier of the backup that was included in the response of the management client when the backup was started (in the message Backup _`backupid`_ started from node_`managementnodeid`_). 3. The management client acknowledges the abort request withAbort of backup_`backupid`_ ordered.
Note
At this point, the management client has not yet received a response from the cluster data nodes to this request, and the backup has not yet actually been aborted. 4. After the backup has been aborted, the management client reports this fact in a manner similar to what is shown here: 

Node 1: Backup 3 started from 5 has been aborted.  
  Error: 1321 - Backup aborted by user request: Permanent error: User defined error  
Node 3: Backup 3 started from 5 has been aborted.  
  Error: 1323 - 1323: Permanent error: Internal error  
Node 2: Backup 3 started from 5 has been aborted.  
  Error: 1323 - 1323: Permanent error: Internal error  
Node 4: Backup 3 started from 5 has been aborted.  
  Error: 1323 - 1323: Permanent error: Internal error

In this example, we have shown sample output for a cluster with 4 data nodes, where the sequence number of the backup to be aborted is 3, and the management node to which the cluster management client is connected has the node ID 5. The first node to complete its part in aborting the backup reports that the reason for the abort was due to a request by the user. (The remaining nodes report that the backup was aborted due to an unspecified internal error.)
Note
There is no guarantee that the cluster nodes respond to anABORT BACKUP command in any particular order.
The Backup _`backupid`_ started from node_`managementnodeid`_ has been aborted messages mean that the backup has been terminated and that all files relating to this backup have been removed from the cluster file system.

It is also possible to abort a backup in progress from a system shell using this command:

$> ndb_mgm -e "ABORT BACKUP backup_id"

Note

If there is no backup having the ID_backupid_ running when anABORT BACKUP is issued, the management client makes no response, nor is it indicated in the cluster log that an invalid abort command was sent.