Medusa the central resource and storage manager in the Acheron Service Framework.

Data backup challenges

In the conventional development process of server or client applications, the developer usually has to think about which hardware platform he is developing on, in which format data is stored and how much hard disk space is available on the server or the connected network drives, or NAS/SAN. Even using the latest technologies, such as virtual hard disks, the initial size of the available hard disk space is limited, i.e. a virtual hard disk can never be larger than the sum of the connected physical hard disks.

During operation, this often results in the need to manually or automatically move or archive data from the hard disks to ensure smooth operation. The development department or the operator in the data center are bound to the existing hardware equipment and have to adapt the software solutions to these specifications, i.e. invest resources and time in costly mechanisms to store the data easily and secure it permanently.

In addition, there is often no standardized procedure for developers that prescribes how data is to be stored in a distributed service landscape, i.e. over time a heterogeneous proliferation of the most diverse data formats and mechanisms grows up in an attempt to secure the accumulating data volumes.

Central resource and memory management for developers

Irrespective of this, non-existent archiving and storage mechanisms are a grievance that cannot be tolerated and can not only cause a company serious financial damage, but also great damage to its image. The development costs and maintenance costs of the existing software landscape increase disproportionately with each proprietary storage interface, as do potential sources of errors due to the increasing complexity of the sum of individual solutions.

Medusa solves this problem very elegantly by providing the developer with a simple tool that eliminates the need to worry about whether the size of the available storage is sufficient or where and how the data is ultimately physically stored.

Medusa provides web service and message based interfaces to not only centrally store data generated in a distributed environment, but also to partially or completely mirror and archive data to other locations according to predefined rules.

Medusa also offers a provisioning mechanism that allows to allocate fixed or variable growing contingents to an application or a customer and to calculate them according to their origin. It goes without saying that an automated retention policy mechanism ensures that data can also "age out" according to individually definable rules, in order to be either deleted or saved somewhere else afterwards.

Medusa not only supports standard network drives and NAS/SAN storage, but also offers a direct connection to Amazon's S3 cloud storage environment. Data can be transferred directly from the local Medusa cache to the cloud and back. This is completely transparent in operation and the management of where the data is and how it is stored is handled by Medusa.

Through a generic plugin mechanism, you can of course also connect other storage cloud services besides Amazon S3 to Medusa.