Notes

Outline

LEA Section 7, Chapter 5: Consolidation Design

Design the Services Areas Design enterprise technology architecture based on IT services area. The IT service areas are derived from the business workload analysis and performance requirements. The IT service areas balance the peak hours. The IT services sites are selected to support the IT services area.

IT service areas example

Select the Services Site Design enterprise technology architecture based on IT services area. The IT service areas are derived from the business workload analysis and performance requirements. The IT service areas balance the peak hours. The IT services sites are selected to support the IT services area.

Select the IT service site

Application and data locality  Application and Data can be located in four areas: the data center, the departmental server,  the local servers and the workstations. The data center is used for mission-critical applications.  These systems use the WAN and LANs to provide access to users. Each office (regional, district, etc.) is being equipped with a LAN, servers, printers, and workstations Another information resource is the LAN file server. the file server appears as an adjunct to local workstation storage. Applications written for the workstation can transparently use the file server as a data file repository or to load applications for local execution. Workstations have local hard drives. Although they can be used for data storage, they must not be depended upon too heavily because there is typically no regular backup of data on a workstation’s local hard disks. If regular file backup is required, the information should be placed on the LAN file server or a local application server.

Locality Decision Tree

Decision Tree Example

Design Factor Based on organization. Some of the decisions made in the tree will require research and analysis to gather data necessary to allow informed decisions to be made. Examples of information needed include the following: The demographics of the users The number of transactions generated per unit time and amount of data transferred per transaction An understanding of the WAN/LAN traffic created by the application (which can also be used to determine geographic location of the application’s server/servers) and the relative costs and efficiencies of communication lines versus more localized servers The impact of the selected middleware solution (that is, use of a transaction processing monitor, CORBA/DCOM) and its effect on network traffic Whether the existing infrastructure can support a fat client or servers can support a large volume of thin clients The complexity of supporting, managing, and maintaining application data if they were more widely distributed Skills required of system administrators and the relative costs of training staff compared to deploying a more centralized approach.

Data placement Factors to be considered when determining data placement within the architecture include the following: Data Volatility – Frequency with which data are updated Data Volume – Amount of data that must be stored Scope of Interest – Community of interest (for example, district, region, or nationwide) and who is the primary INS unit responsible for the data Shared Access – Population of users who must share the data Data Retention – Length of time data must be retained Data Protection – Confidentiality, integrity, and availability Data Accessibility – Speed with which data must be made available (for example, should it be stored online, near-line, or off-line).

Local Information In general, applications with less data and less frequent update requirements are suitable for the local file server. These applications involve documents such as letters, forms, and spreadsheets; discrete data elements such as counts and property; and fairly static data requiring minimal updates.

Share information Shared data access considers which users need concurrent access to the information. Some business processes benefit from sharing data. Data redundancy and reentry are reduced or eliminated when data can be effectively shared. Sharing electronic information includes determining who should have rights to create, read, update, and delete pieces of information. Shared data may be very dynamic, such as the office’s calendar—which may be electronically shared among staff for review and modification—or statistical data that are being updated frequently. Some shared data will be fairly static over its lifetime. Examples of shared static data include application configuration and parameter files. These are used by everyone using the application, but are typically read-only

Data Retention Data retention and data accessibility can be addressed together. Data retention indicates the length of time data must be retained. The longer the data must be retained, the larger the data store grows. In conjunction with data retention requirements, data accessibility addresses the speed with which a data set must be accessed. Requirements can span between subsecond response times to overnight batch staging of data, depending on the business requirements

Storage and Location These retention and accessibility requirements drive the choice of storage media and location. In terms of retrieval speed, retrieval from local magnetic disk storage is quickest. Other storage areas, in order of fastest to slowest retrieval speed, include LAN-attached storage (that is, from the file server) or a computer’s magnetic disk storage, optical storage, and magnetic tape. Magnetic disk storage is typically referred to as online storage. Multiplatter optical storage, such as jukeboxes, or automated magnetic tape library systems are referred to as near-line storage because media can be automatically loaded upon request without human intervention. Magnetic tape or optical platters sitting on a shelf are off-line storage. These must be physically inserted into a drive by a system operator.