cubicweb-s3storage 4.0.1

Summary

A Cubicweb Storage that stores the data on S3.

It is a standard Cubicweb Storage, similar to the BytesFileSystemStorage. For example, if you want to store File.data payloads in S3, you should create an S3Storage instance, typically in a startup hook, as:

from cubicweb.server.hook import Hook
from cubicweb_s3storage.storages import S3Storage

class S3StorageStartupHook(Hook):
    __regid__ = 'mycube.server-startup-hook'
    events = ('server_startup', 'server_maintenance')

    def __call__(self):
        storage = S3Storage('s3_bucket')
        self.repo.system_source.set_storage('File', 'data', storage)

By default, S3 object keys (ie. S3 object identifiers) are random UUID generated at entity creation time. If you want to implement another key generation algorithm, you must write overload S3Storage.new_s3_key() method.

Communications with AWS are handled by boto3.

Configuration

The S3Storage class constructor needs an S3 bucket name.

If you want to put your aws credentials in an other file than the default (~/.aws/credentials), you must set the AWS_CONFIG_FILE environment variable.

If you want to use another profile name than default, set the AWS_PROFILE environment variable.

If you prefer to configure AWS credentials by environment variables (eg. when using docker), you can use the following variables:

• AWS_ACCESS_KEY_ID: the access key for your AWS account,

• AWS_SECRET_ACCESS_KEY: the secret key for your AWS account.

Please read boto3 documentation for more details.

Further configuration

In some cases you might want to use compatible S3 storage which is not on AWS.

For this you have the following variables :

• AWS_S3_ENDPOINT_URL: endpoint url for your alternative S3 storage

This enables the use of a minio server, see their documentation about endpoints.

Note : this type of variable has been requested upstream (and related aws-cli).