A File-Distribution Approach to Achieve High Availability and Confidentiality for Data Storage on Multi-cloud

Abstract

Cloud services are rapidly increasing. We usually use only one cloud service. However, today costs of using cloud service become lower, so we can use multiple cloud services simultaneously. In these environments, there appear possibilities to improve confidentiality and availability by using multiple hetero cloud services. Based on a "secret sharing scheme," a new approach to attain extremely high confidentiality and high availability at reasonable cost is proposed. With this approach, multiple optimal clouds for storage are dynamically determined in accordance with each file and cloud SLA (service level agreement). Furthermore, the reliability of the proposed approach is quantitatively demonstrated by using an evaluation model in the case of homo cloud services.

An Approach to Selecting Cloud Services for Data Storage in Heterogneous-Multicloud Environment with High Availability and Confidentiality

“…They proposed a heuristic model for c configuration for efficient use. Furthe distribution approach using a secret shari proposed and evaluated in a homo-multi-cl [12]. A data management approach in this also proposed [13].…”

“…We explained an evaluation in previous work [12] using some of the metrics in a homogeneous-clouds environment. This paper presents an evaluation in a heterogeneous-clouds environment, which is opposite to the homogeneous-clouds environment.…”

“…We assumed a multi-cloud environm cloud-storage services in this scheme, a services had machine readable SLAs that XML [12], [18]. All SLAs had conditions needed to match.…”

“…We have proposed an approach using scheme [12][13] that introduced a data distr and a data management scheme. When a cloud services, it is distributed using a (k, L, scheme [19].…”

Secret Sharing Scheme for Data Confidentiality Preserving in a Public-Private Hybrid Cloud Storage Approach

“…Avoidance of the over-selection of certain volumes is done using the AN T ESELECT ED queue (line [12][13][14]: once a virtual volume is selected, it is queued in AN T ESELECT ED; a storage volume can be a candidate of a selection only if it has not been used recently in previous selections; when AN T ESELECT ED queue contains all the utilized volumes in storing the file, the queue is emptied. The V UT ILIZAT ION queue manages the set of utilized volumes that stores all chunks of a file (line [15][16][17]: if a selected volume has not been previously utilized in storing the file chunk it is queued in V UT ILIZAT ION; k v is the number of utilized volumes in storing the chunks of a single file. After the destination storage volume has been identified the chunk is sent on the storage provided by the public cloud vendor and the transaction is logged in the Chunk Distribution Dictionary residing on the private cloud infrastructure.…”