Knowledge Base
By: Jaya Wazirani Agarwal
Grid computing is a new IT architecture that leads to more reliable, recoverable and lower cost
enterprise information systems. By efficient use of grid computing, groups of independent, modular
hardware and software components can be connected and rejoined to meet the changing needs of
The motto of grid computing is to resolve some common problems with enterprise IT: the problem of
application clusters that lead to under utilized, dedicated hardware resources; the problem of large,
unwieldy systems that are very expensive to maintain and difficult to change; and the problem of
distributed, fragmented and disintegrated information that cannot be fully exploited by the enterprise
as a whole.
For over a decade Oracle has been providing software that allows organizations to adopt distributed
computing architectures as part of their business transactional applications, content management
applications, and business intelligence applications. Grid or utility Computing is based upon a wellestablished foundation. A distributed computing architecture often utilized in the world of highperformance technical computing. This approach allows organizations to harness together the power
of the newest generation of high-performance and low-cost systems. This paper represents how
Oracle 10g allows organizations to deploy Grid Computing as a foundation for business-oriented
transactional, content management, and business intelligence applications.
The grid computing treats all IT resources as a single resource and exploits the distinct nature
of individual resources within single pool
Grid computing achieves a balance between the benefits of collective resource management and
flexible independent resource control. IT resources managed in a grid include:
• Infrastructure - the hardware and software that create a data storage and program execution
• Applications - the program logic and flow that define specific business processes;
• Information – the meanings inherent in all different types of data used to conduct business.
Knowledge Base
Infrastructure Grid
Infrastructure grid resources include hardware resources such as storage, processors, memory,
and networks, as well as software designed to manage this hardware, such as databases,
storage management, system management, application servers, and operating systems.
Virtualisation and provisioning of infrastructure resources mean pooling resources together and
allocating to the appropriate consumers based on policies. For example, one policy might be to
dedicate enough processing power to a web server that it can always provide sub-second
response time. That rule could be fulfilled in different ways by the provisioning software in order to
balance the requests of all consumers.
Treating infrastructure resources as a single pool and allocating those resources on demand
saves money by eliminating under utilized capacity and redundant capabilities. Managing
hardware and software resources holistically reduces the cost of labour and the opportunity for
human error.
Spreading computing capacity among many different computers and spreading storage capacity
across multiple disks and disk groups removes single points of failure so that if any individual
component fails, the system as a whole remains available. Grid computing also affords the option
to use smaller individual hardware components, such as blade servers and low cost storage,
which enables incremental scaling and reduces the cost of each individual component, giving
companies more flexibility and lower cost.
Infrastructure is the dimension of grid computing that is most familiar and easy to understand, but
the same concepts apply to applications and information.
Storage Virtualisation
Oracle Automatic Storage Management (ASM), a feature of Oracle Database 10g, provides a
virtualisation layer between the database and storage so that multiple disks can be treated as a
single disk group and disks can be dynamically added or removed while keeping databases
online. Existing data will automatically be spread across available disks for performance and
utilization optimisation. In Oracle Database 10g Release 2, ASM supports multiple databases,
which could be at different software version levels, accessing the same storage pool.
Grid Management
Because grid computing pools together multiple servers and disks and allocates them to multiple
purposes, it becomes more important that individual resources are largely self-managing and that
other management functions are centralized.
The Grid Control feature of Oracle Enterprise Manager 10g provides a single console to manage
multiple systems together as a logical group. Grid Control manages provisioning of nodes in the
grid with the appropriate full stack of software and enables configurations and security settings to
be maintained centrally for groups of systems.
Another aspect to grid management is managing user identities in a way that is both highly
secure and easy to maintain. Oracle Identity Management 10g includes an LDAP-compliant
directory with delegated administration and in Release 2, federated identity management, so that
single sign-on capabilities can be securely shared across security domains. Oracle Identity
Management 10g closely adheres to grid principles by utilizing a central point for applications to
authenticate users – the single sign-on server – while distributing control of identities via
delegation and federation to optimise maintainability and overall operation of the system.
Applications Grid
Business Process Management
Knowledge Base
Business Process Execution Language (BPEL) provides a standard for orchestrating processes
into complex business flows in service oriented architecture. The Oracle BPEL Process Manager
10g is the industry’s first native implementation of BPEL for modelling, deploying, and managing
business flows in a standards-compliant way. It comprises an easy-to-use BPEL modeller, a
scalable native BPEL engine, an extensible WSDL binding framework, a monitoring console, and
a set of built-in integration services.
Oracle Business Activity Monitoring 10g (BAM) enhances Oracle’s support for managing
business processes by providing real-time visualization of business activities, alerts and
notifications based on business-level thresholds, metrics and key performance indicators, and
overall monitoring of business events.
Standard Web Services Support
Oracle Application Server 10g provides a cohesive SOA (Service Oriented Architecture) platform
consisting of four main components. First, Oracle Containers for J2EE (OC4J) is a
comprehensive, J2EE-certified service oriented architecture platform to develop and deploy
simple and composite Web Services. Oracle 10g Release 2 includes support for J2EE 1.4, which
includes multiple standards for implementing Web Services with Java.
Second, Oracle JDeveloper is the integrated J2EE development environment for modelling,
developing, debugging, optimising, and deploying Java applications and Web Services.
Third, the Oracle Application Development Framework (ADF), which is embodied by default in
JDeveloper, provides an SOA framework based on the Model-View-Controller design pattern that
will dramatically improve developer productivity.
Finally, Oracle TopLink provides an object-relational mapping solution and middle tier persistence
layer that simplifies how J2EE applications map to and access relational and XML-oriented data.
Oracle Enterprise Manager 10g enhances Oracle’s support for SOA by monitoring and managing
Web Services and any other administrator-defined services, tracking end-to-end performance and
performing root cause analysis of problems encountered.
Information Grid
Data Provisioning
Information starts with data, which must be provisioned wherever consumers need it. For
example, users may be distributed geographically, and fast data access may be essentially
important for these users than access to an identical resource. In these cases, data must be
shared between systems, either in bulk or real time. Oracle’s bulk data movement technologies
include Transportable Tablespaces and Data Pump.
For more fine-grained data sharing, the feature of Oracle Database 10g captures database
transaction changes and propagates them, keeping two or more database copies in sync as
updates are applied. It also unifies traditionally distinct data sharing mechanisms – message
queuing, replication, events, data warehouse loading, notifications, and publish/subscribe – all
into a single technology.
Centralized Data Management
Oracle Database 10g manages all types of structured, semi-structured, and unstructured
information, representing, maintaining and querying each in its own efficient way while providing
common access to all via SQL and XML Query. Along with traditional relational database
structures, Oracle natively implements OLAP cubes, standard XML structures, geographic spatial
data, and unlimited sized file management, which virtualises information representation.
Combining these information types enables connections between disparate types of information
to be made as readily as new connections are made with traditional relational data.
Knowledge Base
Metadata Management
Oracle Warehouse Builder is more than a traditional batch ETL tool for creating warehouses. It
enforces rules to achieve data quality, does fuzzy matching to automatically overcome data
inconsistency, and uses statistical analysis to infer data profiles. With Oracle Database 10g
Release 2, its metadata management capabilities are extended from scheduled data pulls to
handle a transaction-time data push from an Oracle Database implementing the Oracle Streams
Oracle’s series of enterprise data hubs (e.g. Oracle Customer Data Hub) provides real-time
synchronization of operational information sources so that companies can have a single source of
truth while retaining separate systems and separate applications, which may include a
combination of packaged, legacy, and custom applications. In addition to the data cleansing and
scheduling mechanisms, Oracle also provides a well-formed schema, established from years of
experience building enterprise applications, for certain common types of information, such as
customer, financial, and product information.
Metadata Inference
Joining the Oracle 10g software family in 2005 is the new Oracle Enterprise Search product.
Oracle Enterprise Search 10g crawls all information sources in the enterprise, whether public or
secure, including email servers, document management servers, file systems, web sites,
databases, and applications, then returns information from all of the most relevant sources for a
given search query. This crawl and index process uses a series of heuristics specific to each data
source to infer metadata about all enterprise information that is used to return the most relevant
results to any query.
Benefits of Grid Computing
Compared to other models of computing, IT systems designed and implemented in the grid style
deliver a higher quality of service, at a lower cost, with greater flexibility. Higher quality of service
results from having no single point of failure, a powerful security infrastructure, and centralized,
policy-driven management. Lower costs derive from increasing the utilization of resources and
dramatically reducing management and maintenance costs. Rather than dedicating a stack of
software and hardware to a specific task, all resources are pooled and allocated on demand,
which eliminates under-utilised capacity and redundant capabilities. Grid computing also enables
the use of smaller individual hardware components, which reduces the cost of each individual
component and providing more flexibility to devote resources in accordance with changing needs.
The motivation for the development of grid computing is to reduce the need to have dedicated
resources sized for peak capacity. Research has shown that the average CPU usage is only 1520% and storage usage is only 50%. Having dedicated resources for each major application also
means that there can be a large number of systems to maintain. The solution to this is to create a
grid - a pool of low-cost servers + storage that can be allocated to applications to meet peak
Other enhancements for Oracle 10g are aimed at reducing costs and improving the quality of
service by making the database easier to deploy and manage including backup and recovery and
data warehousing enhancements.
1. Ease of Management
In Oracle 10g, a new Automated Storage Management (ASM) sub-system has been introduced.
This feature removes the provision to have a file system or a volume manager for managing the
database files and eliminates the complexity of balancing data across disks and controllers
Knowledge Base
manually. Instead, the ASM stripes data across the disks automatically, thereby maximizing
throughput and reducing the risk of data loss. The ASM also maintains the data striping as disks
are added/removed and re-balances the I/O load dynamically.
Oracle 10g has also been given a self-management infrastructure which captures information
about the running of the database and stores it in the Automatic Workload Repository. This
information is used by tools such as the new Automatic Database Diagnostic Monitor (ADDM)
which analyses data such as the top SQL statements and passes this onto the SQL Tuning
Advisor which provides recommendation for improving performance. These recommendations
can be implemented transparently to the application making it a useful feature for 3rd-party
Other major enhancements in this area include:- enhancing Enterprise Manager to be able to
manage the whole grid and the complete stack of resources and simplifying installation and
configuration by reducing the number of initialisation parameters - for example there are now just
two memory size parameters, one each for the SGA and the PGA.
2. Availability
Real Application Clusters have been enhanced to provide Automatic Service Provisioning servers are automatically allocated to work loads and clients are automatically assigned to the
server with the least load. Also on failure of a server, the surviving servers are automatically reallocated to work loads.
Enhancements in this area also include:
Flashback available at the row, transaction, table or database level;
Recovery area on disk that is maintained automatically by the database and contains
only those blocks changed since the last backup - thereby enabling faster recovery from
media failure;
data guard (standby database) has been enhanced to enable compression and
encryption of log traffic from the master database to the standby system;
Tables can now be re-defined without invalidating stored procedures;
Support for rolling upgrades of the hardware, operating system and the database to
reduce planned down time.
3. Data Warehousing
Data Warehousing enhancements include:
An increase in the size limits of the database to support ultra-large databases totalling
millions of terabytes with ultra-large files of terabytes in size. The 4GB restriction on
LOBs has been raised to 128 terabytes.
Improvements to Real Application Clusters (RAC) enable resources to be allocated
automatically that means that operational data can be used immediately without the need
to copy it to another database.
Enhancements to OLAP analytics, a data-mining GUI and a new SQL model allow query
results to be treated as sets of multi-dimensional arrays on which complex interdependent operations - such as forecasting - can be run without the need to extract data
to spreadsheets or perform complex joins and unions on the data.
Knowledge Base
Where does grid computing fit:
First(Host Based
Second(Remote Access)
Third (Client Server)
Fourth (Mulitier)
Fifth(Grid Computing)
• Dumb Terminal
• Single Server
• Bunch of Complex applications
• Single Server
• Single Client
• Multiple Clients(But Limits)
• Upto Two servers
• Mulitiple Client
• More than two servers
• Virtual environment where all systems are treated as single pool
of resources
• N tier
• Service Oriented Architecture
Entering the Fifth Generation. Grid Computing
Grid Computing is the result of several trends coming together. Some of these are as
High-performance microprocessors have become available, making it possible to deploy
large applications on a number of low-cost systems rather than a single mid-range
New standards for object-to-object communications making it easier to build multivendor,
multiapplication networks.
High-speed networking technology is becoming both less costly and readily available,
offering higher levels of performance when deploying distributed application
As these trends combine, applications are likely to be segmented by function or instance of a
function. This approach will allow each function to be hosted on the most cost-effective platform.
In some cases, both types of segmentation will be used. In the end, an organization’s systems
can be considered a pool of shared resources that adapt automatically to changing conditions
and failures based upon rules of the organization’s choosing.
Grid Resources Work Well Independently and Best Together
By managing any single IT resource – infrastructure, applications, or information – using grid
computing, regardless of how the other resources are treated, enterprises can realize higher
quality, more flexibility, and lower costs. It’s also possible to deploy an applications grid, or an
SOA, without changing the way information is managed or the way hardware is configured.
It’s possible, however, to derive even greater benefit by using grid computing for all resources.
For example, the applications grid becomes even more valuable when you can set policies
regarding resource requirements at the level of individual services and have execution of different
services in the same composite application handled differently by the infrastructure – something
that can only be done by an application grid in combination with an infrastructure grid. In addition,
building an information grid by integrating more information into a single source of truth becomes
Knowledge Base
tenable only when the infrastructure is configured as a grid, so it can scale beyond the boundary
of a single computer.
Oracle’s Vision for Grid Computing
Infrastructure resources managed in a grid will progress to the point that computing and storage
capacity are delivered on demand like a utility. Applications in a grid will advance so that business
and application logic are as massively connected and referenced as static web pages are on the
Internet today, enabling frictionless, automated, global business between trading partners.
Eventually, a global information grid will impart to every bit of digitally-represented information
anywhere the same values we take for granted with relational databases; it will be as if all
information resides in a single virtual database. All inherent relationships between information will
be revealed, and anyone with appropriate authorization will have instantaneous access to all
relevant information regardless of representation, location, or access method.
Grid computing is the next generation model for enterprise computing based on the core tenets of
virtualisation and provisioning of every resource in IT. Grid computing delivers benefits of
increased utilization and greater flexibility for infrastructure, applications, and information
resources. Oracle 10g is the family of software products that supports grid computing and the
software foundation that is enabling many companies and institutions to make the promise of grid
computing a reality.
Download PDF