|
|
 |
 |
>>
Performance
Parameters System |
|
 |
| Performance
Parameters System |
| Customer
and/or End Users |
The
customer is a leading provider of business intelligence
and custom database systems, with several Fortune
500 clients.
|
| The Challenge |
The
company managed a very large number of databases
used for mission critical tasks, run with sophisticated
data manipulation algorithms. The scale of operations
necessitated the utilization of several database
server farms. The entire set up was so huge that
it became almost impossible to gather performance
data from each database server within a viable
timeframe. For this reason, the growth velocity
of the business was slowing down, and management
needed a solution that would accomplish the collection
of performance data at one place, and in the minimum
possible time.
|
| The
Solution |
After
studying the client's existing set up, PM AM
proposed a Database Farm Dashboard as a user
interface, backed up a communication tool to
each database in operation. The idea was that
the system would communicate periodically with
every database among the very large volume of
database farms in operation, and collect data
from its performance log file. This data would
then be aggregated, analyzed and displayed in
a series of meters showing mission-critical
performance parameters. Thus, with the minimum
intervention and elapsed time, performance monitoring
of the entire set of database server farms could
be accomplished.
|
| Technology |
Java,
Web Services support through GLUE technology,
Oracle database, Microsoft SQL server 2000
|
| Resulting
Benefits |
| The
most significant gain to the organization was
the implementation of what was rapidly shaping
up to be an impossible task-in a timely and ongoing
manner. This proved to be an efficient and simple
way to gather and evaluate database server performance
data. As the dashboard now appeared on the concerned
personnel's desktops, showing current and actionable
performance metrics, this led to easier maintenance
of the company's database servers. |
| |
|
|
|