Page 279 - Big data - Concept and application for telecommunications
P. 279
Big data - Concept and application for telecommunications 5
I.6 Data virtualization CSP:SFP
Title Data virtualization by CSP:SFP
Description This scenario shows data virtualizations through the description of read and write data. DSF
service customer #1 writes a file through the DSF service. DSF service provider prepares to
write data because file A, shown in Figure 1, is saved on a single virtual volume. As the
preparation information (e.g., owner, policies and corresponding APIs of data storage, etc.),
data operation metadata for the file A is generated. The file A namely data A, is saved and
managed by the data operation metadata, such as number of data portioning, data location,
etc. In this procedure, the file A was abstracted for data provisioning and management by
policies.
DSF service customer #1 reads a file through the DSF service. DSF service provider gathers data
based on the data operation metadata. DSF service provider aggregates partitions of data and
then data is provided to DSF service customer #1. It is possible for the provided data to be
used again. Therefore, the data is cached and reflected on data operation metadata. For this
reason, DSF service customer #2 quickly reads the cached data. In this procedure, data A was
virtualized because the partitioned data A was aggregated as if originally single data A was and
then was used for multiple DSF service customers. That is, DSF service customer #1 and #2
read same data.
Role/Sub-role DSF service customer (CSC:CSU)
DSF service provider (CSP:SFP)
DSF local storage provider (CSP:CSM)
Figure (optional)
Figure 1 – Writing data by abstracting data
Network and infrastructure 271