Examen 3. Presentacion 4 Flashcards
What is cloud computing?
Cloud Computing (CC) is a model for enabling
ubiquitous, convenient, on-demand network
access to a shared pool of configurable
computing resources (e.g., networks, servers,
storage, applications, and services) that can be
rapidly provisioned and released with minimal
management effort or service provider
interaction
public cloud
when it is made available in a pay-as-you-go
manner to the general public
private cloud
when the cloud infrastructure is
operated solely for a business or an organization
Cloud Computing services
Software-as-a-Service (SaaS), Platform-as-a-Service (PaaS), Infrastructure-as-a-Service
Software-as-a-Service (SaaS)
Applications are accessible from several client devices.
The provider is responsible for the application for example : Email, CRM, Collaborative, ERP.
Platform-as-a-Service (PaaS)
The client is responsible for the end-to-end life cycle in
terms of developing, testing and deploying applications
Providers supplies all the systems (operating systems,
applications, and development environment)
For Example: Application Development, Decision Suppport, Web, Streaming
Infrastructure-as-a-Service (IaaS)
In this type of service the client manages the
storing and development environments for Cloud
Computing application such as the Hadoop
Distributed File System (HDFS) and the MapReduce
development framework.
For example: Caching, Legacy, Netoworking, etc.
NoSQL
Not Only SQL, refers to an
eclectic and increasingly familiar group of nonrelational data management systems; where
databases are not built primarily on tables, and
generally do not use SQL for data manipulation.
NoSQL Focus
NoSQL databases focus on
analytical processing of large scale datasets,
offering increased scalability over commodity
hardware.
Why NoSQL Databases?
1.- The exponential growth of the volume of data
generated by users, systems and sensors.
2.-The increasing interdependency and complexity of
data accelerated by the Internet, Web2.0, social
networks and open and standardized access to
data sources from a large number of different
systems.
The CAP-Theorem
postulates that only two of
the following three different aspects of scaling
out can be achieved fully at the same time
Aspects of scalling
Strong Consistency, High Availability, Partition-tolerance
CAP theorem and NoSQL
Many of the NoSQL databases above all have loosened
up the requirements on Consistency in order to achieve
better Availability and Partitioning (AP).
primary uses of NoSQL Database
- Large-scale data processing (parallel processing over
distributed systems). - Embedded I-R (basic machine-to-machine information
look-up & retrieval). - Exploratory analytics on semi-structured data (expert
level). - Large volume data storage (unstructured, semi-structured,
small-packet structured).
Classification of NoSQL Databases
Key-Value stores, Document databases, WideColumn, Graph databases
Key-Value store
these Data Management Systems (DMS) store items as alphanumeric identifiers (keys) and associated values in
simple, standalone tables (referred to as ―hash
tables‖). The values may be simple text strings or
more complex lists and sets.
Document databases
designed to manage
and store documents which are encoded in a
standard data exchange format such as XML,
JSON or BSON
Key-Value store values
The values may be simple text strings or
more complex lists and sets.
Document Databases Value
he values may be simple text strings or
more complex lists and sets.
Wide-Column
This type of
NoSQL Database employs a distributed,
column-oriented data structure that
accommodates multiple attributes per key
Graph Databases
Graph databases replace
relational tables with structured relational
graphs of interconnected key-value pairings
Hadoop HBase
HBase is a column-oriented
database management system that runs on top
of Hadoop Distributed File System (HDFS).
The reason to store values on a per-column
basis
for specific queries, not all of the values are
needed
HBase Architecture
Consistency, Atomic Read and Write, Sharding, High Availability, Client API, Scalability, Distributed Storage, HDFS/Hadoop integration, Data Replication, Load sharing and Support for Failure, API Support, MapReduce Support, Sorted Row Keys, Real Time Processing of Data,
Consistency
HBAse provides consistent read
and write operations and thus can be used for
high speed requirements. This also helps to
increase the overall throughput of the system.
Atomic Read and Write
Atomic read and write
means that only one process can perform a
given task at a given time. For example when
one process is performing write operation no
other processes can perform the write
operation on that data.
Sharding
HBase offers automatic and manual
splitting of regions. This means that if a region
reaches its threshold size it automatially splits into
smaller sub regions.
High Availability
HBase provides Local Area
Network(LAN) and Wireless Area Network(WAN)
which supports failure recovery. There is a master
server which monitors all the regions and metadata of
the cluster
Client API
HBase offers access through the Java API
which helps to programmatically access HBase.
Scalability
This is one of the important characteristics
of non-relational databases. HBase supports scalability
both in linear and modular form
Distributed Storage
This feature of HBase helps usage
of distributed storage such as HDFS
HDFS/Hadoop integration
HBase can run on top of
various systems such as Hadoop/HDFS
Data Replication
The data in HBase are replicated over
a number of clusters. This helps to recover data in case of
any loss and high availability of data
Load sharing and Support for Failure
HDFS is
internally distributed and supports automatic recovery. As
HBase runs on top of HDFS it is also automatically
recovered.
API Support
HBase supports Java API which makes it
easily available programmatically using java
MapReduce Support
HBase supports map reduce
which helps in parallel processing of data.
Sorted Row Keys
HBase uses keys and stores it in
lexicographical order thus optimizing the requests.
Real Time Processing of Data
HBase performs real
time processing of data and supports block cache and
bloom filters
Denormalization, Duplication, and Intelligent Keys
(DDI)
It is about rethinking how data is stored in
Bigtable-like storage systems, and how to make use of
it in an appropriate way
table row keys
are byte arrays so
almost anything can serve as a row key from
strings to binary representations of longs or
even serialized data structures
Design Aspects
Tables are declared up front at schema definition
time, Rows are lexicographically sorted with the lowest
order appearing first in a table, Columns are grouped into column families, The column family prefix must be composed of printable characters, The qualifying tail, the column family qualifier, can be made of any arbitrary bytes, Column families must be declared up front at schema, A {row, column, version} tuple exactly specifies a cell in HBase.
