Performance Evaluation of Object-based Cloud Storage Services

At the pace with which data is growing today, it is not surprising that the popularity of cloud storage is growing. The fastest growing volumes of archived data are ideal for storage in the cloud, subject to a number of conditions, including the conditions of the economy, the frequency of calls, protection, and availability. But not all cloud storage is the same. One supplier may focus primarily on cost, while the other focuses on availability or performance. No architecture is focused on one thing, and the degree to which the architecture implements a given characteristic determines it's market and target usage, models.
It is difficult to talk about architecture without considering its operating parameters. By this, we mean various characteristics of the architecture, including cost, performance, remote access, etc. Therefore, IT Services Nottingham we first define a set of criteria that measures the characteristics of cloud storage models and then consider some interesting implementations.
First, let's talk about cloud storage architecture in general, to determine the context for the subsequent study of the unique features of this architecture.

Common architecture

Cloud storage architecture is primarily the on-demand delivery of storage resources in a highly scalable and multi-tenant environment. Generally, cloud storage architecture is an external interface that provides an API for accessing drives. In traditional storage systems, this is the SCSI protocol, but new protocols appear in the cloud. Among them, you can find external Web services protocols, file protocols, and even more traditional external interfaces (Internet SCSI, iSCSI, etc.). Behind the external interface is the middleware layer, which we call data storage logic. This level implements a number of functions, such as data replication and data reduction, according to traditional data placement algorithms. Finally, the internal interface organizes the physical storage of data. This may be an internal protocol that implements specific functions, or a traditional server with physical disks.

Controllability

One of the key characteristics of a cloud storage system is its cost. If a customer can buy and manage storage resources locally, rather than rent them in the cloud, the cloud storage market will disappear. Its costs can be divided into two general categories, the cost of the physical storage ecosystem itself and the costs of managing it. The cost of management is hidden but is a long-term component of the total cost. For this reason, cloud storage must be largely self-managing.

Access method

One of the most striking differences between cloud and traditional storage systems is the means of access to them. Most providers offer different access methods, but Web services APIs are generally accepted. Many of them are implemented on the principles of REST, which implies an object-oriented scheme developed on top of HTTP (using HTTP as a transport). A stateless REST API is simple and effective. REST-APIs are implemented by many cloud storage/ managed security services providers, like Amazon Simple Storage Service (Amazon S3) as well as IT Support Derby.

Multi-tenancy

One of the key features of cloud storage architecture is called multi-tenancy. This means that storage is used by many users. Multi-tenancy affects many levels of cloud storage, from the application level, where users are allocated isolated namespaces, to the storage level, where individual users or categories of users can be allocated separate physical drives.

Scalability

Scalability can be viewed from several perspectives, but we are mainly interested in the provision of cloud storage resources on demand. The ability to increase storage (both up and down) means improved cost-effectiveness for the user and increased complexity for the cloud service provider.
Scalability should be provided not only for the storage system (functional scaling) itself but also for its throughput (load scaling). Another key feature of cloud storage is the geographical distribution of data (geographic scalability), which allows data to be located as close as possible to the user thanks to the group of cloud storage centers.

High availability

When a cloud/ managed security services provider stores user data, it should be able to return this data to the user on demand. Given the network prostates, user errors and other circumstances, this condition can be difficult and reliable to determine.
There are interesting new high availability schemes, such as information distribution. Cleversafe, which provides private cloud storage services, uses the Information Dispersal Algorithm (IDA) to increase data availability in the face of physical failures and network outages.

Control

The ability of the client to control and manage how his data is stored and the costs associated with this is important. Numerous cloud/ managed security services providers offer management tools that give users greater control over spending.

Efficiency

Storage efficiency is an important characteristic of cloud storage infrastructure, especially given its focus on overall savings. The next section is specifically devoted to costs, and this characteristic is more related to the efficiency of the use of available resources than to their cost.

Cost of

One of the most notable features of managed security services/cloud storage is the ability to provide savings. These savings on the acquisition of drives, their energy supply, repair, as well as storage management. If we consider cloud storage from this point of view (including SLA and increased storage efficiency), it can be beneficial for certain usage models.

An interesting example of a cloud storage solution is IT Support Birmingham.

Cloud Storage Models

Until now, we have talked mainly about cloud storage providers, but there are cloud models that allow users to maintain control over their data. Cloud storage develops in three directions, one of which allows for the merging of the other two to achieve economic efficiency and security.

Conclusion

Cloud storage is an interesting direction in the development of storage models, which opens up new possibilities for building, accessing and administering storage systems in an enterprise. Although today cloud storage is a predominantly consumer technology, it is rapidly evolving towards the enterprise.



Comments are closed.