Data Storage

Federated ITIL Capacity Management Database

To facilitate the overall Capacity Management process, ITIL recommends a Capacity Management Database (CDB). TeamQuest Performance Software provides a federated CDB that makes it possible to analyze and report on distributed performance data without regard to where it is physically stored. This allows you to keep the optimal amount of performance data distributed in a manner that ensures efficient data collection and storage.

TeamQuest's federated CDB offers two data storage options for IT performance metrics. You can choose to keep comprehensive data in local databases, and you can harvest and aggregate selected data from several local databases into one or more centralized enterprise databases.

Using TeamQuest's IT Resource capability, TeamQuest IT Service Analyzer and TeamQuest IT Service Reporter are capable of seamlessly accessing data in the federated CDB without regard to its physical location. This means you can report and analyze on the performance of IT services that span a number of local and enterprise databases. The TeamQuest federated CDB enables a drill-down analysis across the enterprise, locating the root cause of a performance bottleneck by automatically traversing various performance databases as needed.

 

	Enlarge image TeamQuest IT Service Reporter: I/O activity by IT Resource
More Sample Screens Enterprise-wide drill-down (Example One) - Image 1 Enterprise-wide drill-down (Example One) - Image 2 Enterprise-wide drill-down (Example Two) - Image 1 Enterprise-wide drill-down (Example Two) - Image 2 TeamQuest IT Service Analyzer: Multi-Systems CPU TeamQuest IT Service Analyzer:Network Byte Traffic by Interface

Implement ITIL-recommended Capacity Management Database:

• Manages all IT service and component performance data
• Enables combined use of both local databases and centralized enterprise databases
• Seamlessly integrates data from multiple databases into one federated CDB
• Facilitates enterprise-wide performance analysis and capacity planning

Optimally centralize performance data with enterprise databases:

• Select key performance parameters to be harvested to enterprise databases
• Aggregate those parameters into coarser units of time
• Centralize only what is necessary, minimizing network and storage requirements
• Keep summarized central data for longer retention periods
• Optionally use Oracle for your enterprise database
• Maintain detailed performance data on the system where it is collected

Create multiple enterprise databases to organize data:

• By homogeneous system types within individual enterprise databases
• By clustered systems within individual enterprise databases
• For all systems in a multi-tiered application within individual enterprise databases
• By organizational units within individual enterprise databases
• To maximize the scalability of your performance database

Local Database

For efficiency, all metrics are initially stored locally on each server where data is being collected. This is the most efficient place to keep detailed performance data. You can analyze local data at a fine granularity, if necessary, to identify the cause of bottlenecks or slow response. User-defined retention periods for comprehensive local databases are generally shorter to keep database size more manageable.

Enterprise Database

Having a centralized enterprise database gives you the ability to efficiently analyze the performance of many heterogeneous systems together. By centralizing only key performance parameters and aggregating the data into a coarser time granularity, you can limit the network traffic and storage needed for keeping data for longer retention periods while preserving the ability to do detailed analysis on near-term fine-grain data kept locally on the systems being analyzed.

A TeamQuest enterprise database allows you to centralize as much data as needed to meet your requirements. Data is harvested from multiple systems using industry-standard HTTP protocol. Harvesting occurs automatically according to user-defined specifications - you control what data is harvested, how it is summarized, and how often it gets harvested.

Oracle Enterprise Database

Your TeamQuest enterprise databases can optionally be implemented using Oracle rather than TeamQuest's proprietary performance database technology. Storing your performance data in an Oracle database opens your enterprise performance data for access, analysis, and reporting using your own custom tools. You can also use third-party reporting tools, such as Crystal Reports, to integrate and report information from a number of sources, including your Oracle enterprise performance database.

Real-world Uses

You can establish a hierarchy of performance databases to suit your needs, giving you a great deal of flexibility in analyzing service performance.

Local databases can be used for detailed analysis on each system, one or more enterprise databases for summary reports, and another enterprise database to store Key Performance Indicators (KPIs) for capacity planning purposes.

The diagram above shows one possible use of TeamQuest's enterprise database capability. In this case performance data is organized by operating system platform, allowing performance data from all systems running a particular OS to be summarized together in a platform-specific database.

An enterprise database is also useful for efficiently analyzing IT service performance across multiple tiers, determining how a workload is behaving across multiple servers, or quickly finding which system is receiving the most traffic. You may also want to summarize information such as the number of active users you have on different systems.

You can think of the enterprise database as a project-oriented or service-oriented database. It lets you analyze workloads or activities that span multiple systems. A server can belong to any number of enterprise databases, and you can have any number of enterprise databases.