6.1 Data Management Frameworks
The Galaxy platform offers a number of data management
frameworks, geared to managing large volumes of data
and mining them for useful information.
Galaxy runs as a clustered web server farm and is
accessed through a load balancer. No data is stored in
the application tier hence more servers can be added to
the farm on the fy, in order to achieve greater scalability.
The data tier consists of multiple replicated databases
that are accessed in parallel for different functionalities.
Every customer’s data is logically separated by default and
physical separation can be done based on the customer
requirement and service agreement. As the number of
customers increase, the number of database servers can
be increased. However, from an end user’s perspective,
Galaxy will remain as one logical instance with a single
web URL.
Galaxy addresses the problem of application scalability by
allowing the creation of templates for all of its resources
so that they can be reused for subsequent projects. The
upcoming event management system attempts to choose
operators/analysts (users) and distribute user tasks
based on availability. Instead of dedicating specifc sets
of facilities to different users, each event is dedicated to
a user based on profle/skill and availability, with overall
visibility for supervisory users. Through this technique,
user effciency is maximized and balanced automatically.
The Point Metering Framework is used to visually
formulate equations used in various engineering
calculations. It has the fexibility to write performance
calculation logic for almost every piece of equipment
mapped with the application. It can use both automated
and manual data as input to the equations, which can
be defned once and applied to multiple sets of input
parameters. Equations can be nested inside equations
and there is support for both Boolean and numeric
operations. Custom operators can be developed and
included in the framework. Equations can be saved as
“virtual meters” to be calculated periodically.
The Equation Builder Framework is able to simulate
meters with a combination of manual entry and automated
processes and can integrate real-time data points and
time-series data of their history. It can create defned
time-periods for time-sensitive metering, and calculated
meters and can manage the rollover and warehousing of
old meter data.
The Data Mining tools can be used to identify patterns in
data. They can apply a rules based analysis and a wide
range of techniques including a number of different types
of regression, neural networks and clustering to predict
future trends. They offer a wide range of data visualization
methods to aid decision making. There are modules for
fraud detection and revenue assurance.
The Graph Designer Framework allows users to design
graphs of defned variables and time series using a simple
drag-and-drop technique. The framework can support
multi-axis graphs and has pie, bar, line, area, scatter,
heat map, and bubble charts pre-confgured. Default
charts are available for quick chart creation, while custom
GALAXY OPEN PLATFORM FOR M2M ENTERPRISE CITY MANAGEMENT - 34