Storage Provisioning and Networking - Deploying and Managing a Cloud Infrastructure - page 256

Data deduplication is another technique for eliminating duplicate copies of data using
data compression. This technique is applied to improve storage utilization and reduce
network data transfer latencies. Unique chunks of data or byte patterns are identified and
stored during a process of analysis. Subsequent chunks of data are compared to the stored
chunks. Whenever a match is found, the redundant chunk is replaced with a reference that
points to the stored chunk. This technique reduces the amount of data stored considerably
for systems where the same byte pattern for a considerable chunk size of data might occur
multiple times. Examples include machine-generated log data from network hardware.
The technique is different from standard file compression tools that do string matching
inside individual files. Deduplication techniques focus on chunks of data such as entire
files or collection of files to find identical copies and store only one copy.
A much-overlooked approach to adding more capacity is to “manage” the data. A
large chunk of data is not accessed on a regular basis and is a good candidate for archival.
Moreover, careful planning for redundancy level and data backups is useful in reclaiming
large amounts of space. The key challenge here is to find which data can be archived and
compressed without being required and decompressed regularly.

Network Configurations

Network Address Translation (NAT) is a network protocol that allows multiple internal
devices (inside a private or local network) to connect to an external network (such as the
Internet) using the same public IPv4 address. The idea is to modify the IP header as the
packet leaves or arrives at the border gateway or router. IP-masquerading through NAT
was originally designed to conserve IPv4 addresses and delay exhaustion. The technique
makes sure hosts that do not require support for inbound connections do not consume
public IPv4 addresses.
Many different configurations can be created with NAT; some of the more common ones
are described in Table 9.1.

TABLE 9.1 NAT configurations

Configuration Features

One-to-one NAT Internal address (iAddr:iPort) is mapped to an external address

(eAddr:ePort).
Internal host can send packets to external host using iAddr:iPort.
External host can send packets to internal host using eAddr:ePort.

Address
restricted NAT

Internal address (iAddr:iPort) is mapped to an external address
(eAddr:ePort).
Internal host can send packets to external host (hAddr:any) using
iAddr:iPort.
External host (hAddr:any) can send packets to internal host using
eAddr:ePort, if and only if the internal host previously sent a packet
to hAddr:any. Any means this scheme is not port bound.