Your company here — click to reach over 10,000 unique daily visitors

barman-cloud-backup - Man Page

Backup a PostgreSQL instance and stores it in the Cloud




This script can be used to perform a backup of a local PostgreSQL instance and ship the resulting tarball(s) to the Cloud. Currently AWS S3, Azure Blob Storage and Google Cloud Storage are supported.

It requires read access to PGDATA and tablespaces (normally run as postgres user). It can also be used as a hook script on a barman server, in which case it requires read access to the directory where barman backups are stored.

If the arguments prefixed with --snapshot- are used, and snapshots are supported for the selected cloud provider, then the backup will be performed using snapshots of the disks specified using --snapshot-disk arguments. The backup label and backup metadata will be uploaded to the cloud object store.

This script and Barman are administration tools for disaster recovery of PostgreSQL servers written in Python and maintained by EnterpriseDB.

IMPORTANT: the Cloud upload process may fail if any file with a size greater than the configured --max-archive-size is present either in the data directory or in any tablespaces. However, PostgreSQL creates files with a maximum size of 1GB, and that size is always allowed, regardless of the max-archive-size parameter.


usage: barman-cloud-backup [-V] [--help] [-v | -q] [-t]
                           [--cloud-provider {aws-s3,azure-blob-storage,google-cloud-storage}]
                           [--endpoint-url ENDPOINT_URL] [-P AWS_PROFILE]
                           [--profile AWS_PROFILE]
                           [--read-timeout READ_TIMEOUT]
                           [--azure-credential {azure-cli,managed-identity}]
                           [-z | -j | --snappy] [-h HOST] [-p PORT] [-U USER]
                           [--immediate-checkpoint] [-J JOBS]
                           [-S MAX_ARCHIVE_SIZE]
                           [--min-chunk-size MIN_CHUNK_SIZE]
                           [--max-bandwidth MAX_BANDWIDTH] [-d DBNAME]
                           [-n BACKUP_NAME]
                           [--snapshot-instance SNAPSHOT_INSTANCE]
                           [--snapshot-disk NAME] [--snapshot-zone GCP_ZONE]
                           [--snapshot-gcp-project GCP_PROJECT]
                           [--gcp-project GCP_PROJECT]
                           [--kms-key-name KMS_KEY_NAME] [--gcp-zone GCP_ZONE]
                           [--tags [TAGS [TAGS ...]]] [-e {AES256,aws:kms}]
                           [--sse-kms-key-id SSE_KMS_KEY_ID]
                           [--aws-region AWS_REGION]
                           [--encryption-scope ENCRYPTION_SCOPE]
                           [--azure-subscription-id AZURE_SUBSCRIPTION_ID]
                           [--azure-resource-group AZURE_RESOURCE_GROUP]
                           destination_url server_name

This script can be used to perform a backup of a local PostgreSQL instance and
ship the resulting tarball(s) to the Cloud. Currently AWS S3, Azure Blob
Storage and Google Cloud Storage are supported.

positional arguments:
  destination_url       URL of the cloud destination, such as a bucket in AWS
                        S3. For example: `s3://bucket/path/to/folder`.
  server_name           the name of the server as configured in Barman.

optional arguments:
  -V, --version         show program's version number and exit
  --help                show this help message and exit
  -v, --verbose         increase output verbosity (e.g., -vv is more than -v)
  -q, --quiet           decrease output verbosity (e.g., -qq is less than -q)
  -t, --test            Test cloud connectivity and exit
  --cloud-provider {aws-s3,azure-blob-storage,google-cloud-storage}
                        The cloud provider to use as a storage backend
  -z, --gzip            gzip-compress the backup while uploading to the cloud
  -j, --bzip2           bzip2-compress the backup while uploading to the cloud
  --snappy              snappy-compress the backup while uploading to the cloud
  -h HOST, --host HOST  host or Unix socket for PostgreSQL connection
                        (default: libpq settings)
  -p PORT, --port PORT  port for PostgreSQL connection (default: libpq
  -U USER, --user USER  user name for PostgreSQL connection (default: libpq
                        forces the initial checkpoint to be done as quickly as
  -J JOBS, --jobs JOBS  number of subprocesses to upload data to cloud storage
                        (default: 2)
  -S MAX_ARCHIVE_SIZE, --max-archive-size MAX_ARCHIVE_SIZE
                        maximum size of an archive when uploading to cloud
                        storage (default: 100GB)
  --min-chunk-size MIN_CHUNK_SIZE
                        minimum size of an individual chunk when uploading to
                        cloud storage (default: 5MB for aws-s3, 64KB for
                        azure-blob-storage, not applicable for google-cloud-
  --max-bandwidth MAX_BANDWIDTH
                        the maximum amount of data to be uploaded per second
                        when backing up to either AWS S3 or Azure Blob Storage
                        (default: no limit)
  -d DBNAME, --dbname DBNAME
                        Database name or conninfo string for Postgres
                        connection (default: postgres)
                        a name which can be used to reference this backup in
                        commands such as barman-cloud-restore and barman-
  --snapshot-instance SNAPSHOT_INSTANCE
                        Instance where the disks to be backed up as snapshots
                        are attached
  --snapshot-disk NAME  Name of a disk from which snapshots should be taken
  --snapshot-zone GCP_ZONE
                        Zone of the disks from which snapshots should be taken
                        (deprecated: replaced by --gcp-zone)
  --tags [TAGS [TAGS ...]]
                        Tags to be added to all uploaded files in cloud

Extra options for the aws-s3 cloud provider:
  --endpoint-url ENDPOINT_URL
                        Override default S3 endpoint URL with the given one
  -P AWS_PROFILE, --aws-profile AWS_PROFILE
                        profile name (e.g. INI section in AWS credentials
  --profile AWS_PROFILE
                        profile name (deprecated: replaced by --aws-profile)
  --read-timeout READ_TIMEOUT
                        the time in seconds until a timeout is raised when
                        waiting to read from a connection (defaults to 60
  -e {AES256,aws:kms}, --encryption {AES256,aws:kms}
                        The encryption algorithm used when storing the
                        uploaded data in S3. Allowed values:
  --sse-kms-key-id SSE_KMS_KEY_ID
                        The AWS KMS key ID that should be used for encrypting
                        the uploaded data in S3. Can be specified using the
                        key ID on its own or using the full ARN for the key.
                        Only allowed if `-e/--encryption` is set to `aws:kms`.
  --aws-region AWS_REGION
                        The name of the AWS region containing the EC2 VM and
                        storage volumes defined by the --snapshot-instance and
                        --snapshot-disk arguments.

Extra options for the azure-blob-storage cloud provider:
  --azure-credential {azure-cli,managed-identity}, --credential {azure-cli,managed-identity}
                        Optionally specify the type of credential to use when
                        authenticating with Azure. If omitted then Azure Blob
                        Storage credentials will be obtained from the
                        environment and the default Azure authentication flow
                        will be used for authenticating with all other Azure
                        services. If no credentials can be found in the
                        environment then the default Azure authentication flow
                        will also be used for Azure Blob Storage.
  --encryption-scope ENCRYPTION_SCOPE
                        The name of an encryption scope defined in the Azure
                        Blob Storage service which is to be used to encrypt
                        the data in Azure
  --azure-subscription-id AZURE_SUBSCRIPTION_ID
                        The ID of the Azure subscription which owns the
                        instance and storage volumes defined by the
                        --snapshot-instance and --snapshot-disk arguments.
  --azure-resource-group AZURE_RESOURCE_GROUP
                        The name of the Azure resource group to which the
                        compute instance and disks defined by the --snapshot-
                        instance and --snapshot-disk arguments belong.

Extra options for google-cloud-storage cloud provider:
  --snapshot-gcp-project GCP_PROJECT
                        GCP project under which disk snapshots should be
                        stored (deprecated: replaced by --gcp-project)
  --gcp-project GCP_PROJECT
                        GCP project under which disk snapshots should be
  --kms-key-name KMS_KEY_NAME
                        The name of the GCP KMS key which should be used for
                        encrypting the uploaded data in GCS.
  --gcp-zone GCP_ZONE   Zone of the disks from which snapshots should be taken


For Boto:

For AWS:

For Azure Blob Storage:

For libpq settings information:

For Google Cloud Storage: * Credentials: https://cloud.google.com/docs/authentication/getting-started#setting_the_environment_variable

Only authentication with GOOGLE_APPLICATION_CREDENTIALS env is supported at the moment.


If using --cloud-provider=aws-s3:

If using --cloud-provider=azure-blob-storage:

If using --cloud-provider=google-cloud-storage * google-cloud-storage

If using --cloud-provider=google-cloud-storage with snapshot backups

Exit Status




The backup was not successful


The connection to the cloud provider failed


There was an error in the command input

Other non-zero codes


See Also

This script can be used in conjunction with post_backup_script or post_backup_retry_script to relay barman backups to cloud storage as follows:

post_backup_retry_script = 'barman-cloud-backup [*OPTIONS*] *DESTINATION_URL* ${BARMAN_SERVER}'

When running as a hook script, barman-cloud-backup will read the location of the backup directory and the backup ID from BACKUP_DIR and BACKUP_ID environment variables set by barman.


Barman has been extensively tested, and is currently being used in several production environments. However, we cannot exclude the presence of bugs.

Any bug can be reported via the GitHub issue tracker.



Barman is the property of EnterpriseDB UK Limited and its code is distributed under GNU General Public License v3.

© Copyright EnterpriseDB UK Limited 2011-2023


EnterpriseDB <https://www.enterprisedb.com>.


June 12, 2024 Barman User manuals Version 3.10.1