HiveCreator Features
HiveCreator has a number of features that make it unique...
Application FabricTM
An application fabric is a software layer which insulates the
application from the operating system and the hardware upon which it runs.
The application fabric starts from
the premise that the application is the only thing that matters. The
application developer should be able to develop to a single, logical
entity, regardless of the physical implementation and configuration of
the system.
As a result, a Hive
takes over the responsibility for...
- Managing things like fail-over and
fault tolerance.
- Monitoring the execution of tasks
and recovering those that fail.
- Maintaining the configuration of
each Worker.
This has three benefits.
First, it means developers do not have to understand, manage, or
maintain the physical environment.
Instead, they can focus on the thing that really matters: the
application. Second, the reduced complexity of the programming model
means that organizations do not have to hire the most experienced (and
expensive) developers. Third, it allows organizations to use commodity
computers to perform critical functions.
Assimilation
Assimilation is the process of installing a Worker’s
operating system, drivers, or Hive system files. The Hive does this
through the use of an external PXE server.
To assimilate a computer into the
Hive, the administrator installs the desired operating system,
drivers, and Hive system files on one Worker Machine, and creates an
image of that computer. Then the administrator simply connects a
computer to the network. The Hive detects the new computer and
installs the image on the computer.
Hive configuration files and applications are
spread throughout the Hive in a different way.
To add a new application or update a
configuration file or application, the administrator deploys the
application package to one of the Worker Machines and the Hive
automatically spreads the application to all the Worker Machines in
the Hive.
A Hive can quickly and
easily scale to hundreds or thousands of nodes because new Workers can
be quickly and easily added to the pool of available resources.
Collective Intelligence
The collective intelligence
refers to the fact that the intelligence of a Hive is distributed
throughout every Worker instead of being centralized or decentralized.
As a result, a Hive has no single point of failure and can survive the
failure of any number of individual Workers.
The collective intelligence is an inherent property
of a Hive. The application developer does not need to manage or understand this
ability.
Platform Agile
Unlike some solutions which restrict
your choices, HiveCreator allows you to develop using the
hardware, operating systems, and programming language of your choice.
HiveCreator can run on a wide range
of popular operating systems, including Windows® XP, Windows® 2000,
FreeBSD 4.7, and Linux 2.4.
HiveCreator provides Windows C/C++,
UNIX C/C++, and Java® API's. What's more, the HiveCreator API will be
expanded over time to support additional programming languages.
TimeSync
Timesync is the Hive's sense of time. In the
application, developers specify how long a task should take to
process. This provides the Hive with the knowledge of how much time it
takes to process a request. This allows the Hive to return a result
within a predictable amount of time.
As a result, a Hive is able to perform time-definite
processing and is suited for transaction processing.
|
