AWS Backup Construct Library

AWS Backup is a fully managed backup service that makes it easy to centralize and automate the backup of data across AWS services in the cloud and on premises. Using AWS Backup, you can configure backup policies and monitor backup activity for your AWS resources in one place.

Backup plan and selection

In AWS Backup, a backup plan is a policy expression that defines when and how you want to back up your AWS resources, such as Amazon DynamoDB tables or Amazon Elastic File System (Amazon EFS) file systems. You can assign resources to backup plans, and AWS Backup automatically backs up and retains backups for those resources according to the backup plan. You can create multiple backup plans if you have workloads with different backup requirements.

This module provides ready-made backup plans (similar to the console experience):

# Daily, weekly and monthly with 5 year retention
plan = backup.BackupPlan.daily_weekly_monthly5_year_retention(self, "Plan")

Assigning resources to a plan can be done with addSelection():

# plan: backup.BackupPlan
# vpc: ec2.Vpc

my_table = dynamodb.Table.from_table_name(self, "Table", "myTableName")
my_database_instance = rds.DatabaseInstance(self, "DatabaseInstance",
    engine=rds.DatabaseInstanceEngine.mysql(version=rds.MysqlEngineVersion.VER_8_0_26),
    vpc=vpc
)
my_database_cluster = rds.DatabaseCluster(self, "DatabaseCluster",
    engine=rds.DatabaseClusterEngine.aurora_mysql(version=rds.AuroraMysqlEngineVersion.VER_2_08_1),
    credentials=rds.Credentials.from_generated_secret("clusteradmin"),
    instance_props=rds.InstanceProps(
        vpc=vpc
    )
)
my_serverless_cluster = rds.ServerlessCluster(self, "ServerlessCluster",
    engine=rds.DatabaseClusterEngine.AURORA_POSTGRESQL,
    parameter_group=rds.ParameterGroup.from_parameter_group_name(self, "ParameterGroup", "default.aurora-postgresql10"),
    vpc=vpc
)
my_cool_construct = Construct(self, "MyCoolConstruct")

plan.add_selection("Selection",
    resources=[
        backup.BackupResource.from_dynamo_db_table(my_table),  # A DynamoDB table
        backup.BackupResource.from_rds_database_instance(my_database_instance),  # A RDS instance
        backup.BackupResource.from_rds_database_cluster(my_database_cluster),  # A RDS database cluster
        backup.BackupResource.from_rds_serverless_cluster(my_serverless_cluster),  # An Aurora Serverless cluster
        backup.BackupResource.from_tag("stage", "prod"),  # All resources that are tagged stage=prod in the region/account
        backup.BackupResource.from_construct(my_cool_construct)
    ]
)

If not specified, a new IAM role with a managed policy for backup will be created for the selection. The BackupSelection implements IGrantable.

To add rules to a plan, use addRule():

# plan: backup.BackupPlan

plan.add_rule(backup.BackupPlanRule(
    completion_window=Duration.hours(2),
    start_window=Duration.hours(1),
    schedule_expression=events.Schedule.cron( # Only cron expressions are supported
        day="15",
        hour="3",
        minute="30"),
    move_to_cold_storage_after=Duration.days(30)
))

Continuous backup and point-in-time restores (PITR) can be configured. Property deleteAfter defines the retention period for the backup. It is mandatory if PITR is enabled. If no value is specified, the retention period is set to 35 days which is the maximum retention period supported by PITR. Property moveToColdStorageAfter must not be specified because PITR does not support this option. This example defines an AWS Backup rule with PITR and a retention period set to 14 days:

# plan: backup.BackupPlan

plan.add_rule(backup.BackupPlanRule(
    enable_continuous_backup=True,
    delete_after=Duration.days(14)
))

Rules can also specify to copy recovery points to another Backup Vault using copyActions. Copied recovery points can optionally have moveToColdStorageAfter and deleteAfter configured.

# plan: backup.BackupPlan
# secondary_vault: backup.BackupVault

plan.add_rule(backup.BackupPlanRule(
    copy_actions=[backup.BackupPlanCopyActionProps(
        destination_backup_vault=secondary_vault,
        move_to_cold_storage_after=Duration.days(30),
        delete_after=Duration.days(120)
    )]
))

Ready-made rules are also available:

# plan: backup.BackupPlan

plan.add_rule(backup.BackupPlanRule.daily())
plan.add_rule(backup.BackupPlanRule.weekly())

By default a new vault is created when creating a plan. It is also possible to specify a vault either at the plan level or at the rule level.

my_vault = backup.BackupVault.from_backup_vault_name(self, "Vault1", "myVault")
other_vault = backup.BackupVault.from_backup_vault_name(self, "Vault2", "otherVault")

plan = backup.BackupPlan.daily35_day_retention(self, "Plan", my_vault) # Use `myVault` for all plan rules
plan.add_rule(backup.BackupPlanRule.monthly1_year(other_vault))

You can backup VSS-enabled Windows applications running on Amazon EC2 instances by setting the windowsVss parameter to true. If the application has VSS writer registered with Windows VSS, then AWS Backup creates a snapshot that will be consistent for that application.

plan = backup.BackupPlan(self, "Plan",
    windows_vss=True
)

Backup vault

In AWS Backup, a backup vault is a container that you organize your backups in. You can use backup vaults to set the AWS Key Management Service (AWS KMS) encryption key that is used to encrypt backups in the backup vault and to control access to the backups in the backup vault. If you require different encryption keys or access policies for different groups of backups, you can optionally create multiple backup vaults.

my_key = kms.Key.from_key_arn(self, "MyKey", "aaa")
my_topic = sns.Topic.from_topic_arn(self, "MyTopic", "bbb")

vault = backup.BackupVault(self, "Vault",
    encryption_key=my_key,  # Custom encryption key
    notification_topic=my_topic
)

A vault has a default RemovalPolicy set to RETAIN. Note that removing a vault that contains recovery points will fail.

You can assign policies to backup vaults and the resources they contain. Assigning policies allows you to do things like grant access to users to create backup plans and on-demand backups, but limit their ability to delete recovery points after they’re created.

Use the accessPolicy property to create a backup vault policy:

vault = backup.BackupVault(self, "Vault",
    access_policy=iam.PolicyDocument(
        statements=[
            iam.PolicyStatement(
                effect=iam.Effect.DENY,
                principals=[iam.AnyPrincipal()],
                actions=["backup:DeleteRecoveryPoint"],
                resources=["*"],
                conditions={
                    "StringNotLike": {
                        "aws:userId": ["user1", "user2"
                        ]
                    }
                }
            )
        ]
    )
)

Alternativately statements can be added to the vault policy using addToAccessPolicy().

Use the blockRecoveryPointDeletion property or the blockRecoveryPointDeletion() method to add a statement to the vault access policy that prevents recovery point deletions in your vault:

# backup_vault: backup.BackupVault
backup.BackupVault(self, "Vault",
    block_recovery_point_deletion=True
)
backup_vault.block_recovery_point_deletion()

By default access is not restricted.

Use the lockConfiguration property to enable AWS Backup Vault Lock:

# Example automatically generated from non-compiling source. May contain errors.
BackupVault(stack, "Vault",
    lock_configuration={
        "min_retention": Duration.days(30)
    }
)

Importing existing backup vault

To import an existing backup vault into your CDK application, use the BackupVault.fromBackupVaultArn or BackupVault.fromBackupVaultName static method. Here is an example of giving an IAM Role permission to start a backup job:

imported_vault = backup.BackupVault.from_backup_vault_name(self, "Vault", "myVaultName")

role = iam.Role(self, "Access Role", assumed_by=iam.ServicePrincipal("lambda.amazonaws.com"))

imported_vault.grant(role, "backup:StartBackupJob")