Читать книгу IT Architecture from A to Z: Theoretical basis. First Edition - Vadim Aldzhanov - Страница 19

Architecture choice strategy

As a solution architecture, one can consider the following:

• On premise infrastructure;

• Cloud based infrastructure;

• Hybrid architecture.

“On-premise” implies that IT assets are physically located within the organization. Benefits:

• IT infrastructure is located within the organization and controlled by organization’s IT employees.

• Relatively low operative expenditures of IT infrastructure.

• Solution autonomy and higher security.

Deficiencies:

• Relatively high capital expenditure in IT infrastructure.

• The introduction of new services, or surges of existing services are difficult to plan.

• The need to hire IT specialists to maintain physical servers, network equipment, and so on) leads to additional costs.

• Indirect costs for IT engineering systems.

“Cloud based” implies that IT assets are located in the “cloud”.

Benefits:

• Relatively low capital expenditure in IT infrastructure.

• High level of planning of maintenance costs for IT infrastructure.

• Good communication, linear relationship between the resources used and their cost.

• Easiness to implement new solutions and expand existing IT services.

• No need for additional staff.

• No need for indirect engineering costs.

Deficiencies:

• Relatively high operative expenditures of IT infrastructure.

• Higher requirements for Internet bandwidth and backup channels.

“Hybrid” architecture provides combined solutions.

Benefits: The benefits of the first two solutions are used.

Deficiencies: A higher capital and operative (CAPEX and OPEX).

Recommendations for selection:

“Cloud based” solution is more preferable for initial projects, small organizations, organizations with developed geography and so on. “On-site” solution is more preferable for well-developed organizations, financial institutions, organizations where access to the Internet is not a business requirement. “Hybrid” solutions can supplement the IT infrastructure and replace some components of the IT architecture.

Infrastructure platform choice strategy

If you choose “On-premises” as deployment infrastructure, there are the following options:

• Physical servers as a foundation;

• Virtualization platform as a basis;

• Mixed infrastructure with no dominance.

“Physical servers” implies that IT services are located on physical servers. Benefits:

• Relatively low capital expenditure of IT services for small solutions;

• The resources of the physical server are fully allocated to the tasks of a particular service.

Deficiencies:

• The complexity of maintenance as the infrastructure grows;

• Deployment and recovery speed.

• Sub-optimal use of computing resources.

“Virtualization Platform” implies that IT services are located on the virtualization platform as virtual machines. Benefits:

• Relatively low operative expenditures of IT infrastructure.

• The platform is de facto standards for the deployment of “On-premises” solutions.

Deficiencies: Relatively high capital expenditure in IT infrastructure.

Recommendations for selection:

“Physical servers” solution is more preferable for small, isolated or remote IT services, or high volume systems. For all other cases, it is better to use a virtualization platform.

Choice strategy for data storage system

If you choose “On-premises” as deployment infrastructure, there are the following options:

• Physical servers and Direct-Attached Storage (DAS);

• Centralized Data Storage System, i.e. Network Attached Storage (NAS) and Storage Area Networks (SAN);

• Distributed Data Storage System, i.e. Digital Data Storage (DDS) and Software-Defined Storage (SDS).

Physical servers implies that the data is located on physical servers.

Benefits:

• Relatively low capital expenditure of IT services for small solutions.

• The resources of the physical server are fully allocated to the tasks of a particular service.

Deficiencies:

• The complexity of maintenance as the infrastructure grows.

• Deployment and recovery speed.

• Sub-optimal use of computing resources.

Centralized Data Storage System (NAS, SAN) implies that the data are located on a single data storage system partially or completely. DSS is a single complex consisting of controllers and disk storage racks.

Benefits: Relatively low operative expenditures of IT infrastructure.

Deficiencies: Relatively high capital expenditure of IT infrastructure.

Distributed Data Storage System (DDS, SDS)” implies that the data is distributed between different physical servers. A storage system is a complex distributed within the network and consisting of data controllers and disk storages.

Recommendations for selection:

“Physical servers” solution is more preferable for small, isolated or remote IT services, or high volume systems. For all other cases, it is better to use a virtualization platform.

Strategy of choosing the manufacturer

The strategy for choosing software and hardware determines the approach to the choice of manufacturer, standardization, and so on.

The following options can be considered:

• Use of a limited list of manufacturers;

• Use random list of manufacturers.

Using a specific manufacturer for each category of IT assets implies using hardware and software standards within the organization.

Benefits:

• Implementing IT asset standards simplifies providing an organization staff with IT assets;

• It facilitates the implementation and maintenance of IT infrastructure, and increases the level of the organization’s information security.

Deficiencies: Relatively high cost and dependence on the manufacturer/supplier.

Using a random manufacturer implies use of recommendations instead of hardware and software standards within the organization.

Benefits: Relatively low cost and promptness in procurement of IT assets.

Deficiencies:

• The process of providing an organization staff with IT assets is more sophisticated and takes longer;

• It complicates the implementation and maintenance of IT infrastructure, and

• reduces the level of the organization’s information security.

Recommendations for selection:

The first licensing option is recommended is recommended to organizations with close integration and dependence of business and Information Technology or a large number of employees.

Licensing strategy

The licensing strategy determines the approach to licensing methods. The following options can be considered:

Using an “Enterprise” license agreement with the annual software update. Licensing is an ongoing IT process.

Benefits:

• Greater flexibility of licensing.

• Maintaining the high level of IT infrastructure and information security due to using updated versions of systems and solutions.

• Systems are upgraded smoothly, with no surges in requirements of IT, human resources or time.

Deficiencies: Relatively high cost.

Purchasing commercial off-the-shell retail versions with no software update. Licensing “upon request”.

Benefits: A relatively cheaper option.

Deficiencies:

• Less freedom in licensing.

• The level of maintaining IT infrastructure and information security is reduced due to using not the most updated versions of systems and solutions.

• Systems are upgraded in jumps (every three, four years), and require additional IT resources, people, and time.

Recommendations for selection:

The first licensing option is recommended to organizations with close integration and dependence of business and Information Technologies.

The strategy of building engineering systems

“On premise” and “Hybrid” solutions demand to determine the engineering system requirements. The engineering systems requirements for may include:

• Physical requirements for the premises of the data center, server room, data cabinets etc.

• Requirements for a Structured Cabling System (SCS).

• Requirements for redundancy, reservation and so on.

Testing strategy

Testing is one of the important elements when building an IT architecture. Testing questions have to be identified at the design stage. There are the following platforms:

•“Test or Development” is a platform with individual IT services deployed which is used to develop services or adjust IT service elements.

•“Pre-production” is a smaller “Production” platform containing all the components of the IT architecture. It is used to test the interaction of IT services and allows emulating troubleshooting, estimating performance, and so on.

•“Production” is a platform with detailed computing resources providing IT services.

Recommendations for selection:

Various solutions can be implemented depending on the business requirements and the organization’s financial capacity. A detailed description is further discussed in this guide.

Redundancy strategy

Defining the redundancy level of IT infrastructure components indicates the requirements for duplication of components. It can be reviewed as follows:

• Components of engineering systems (channels and communication cables, switching racks, etc.). Reservations (duplication or redundancy) at the level of critical elements, such as:

• Channels and communication cables (two or more in different stacks),

• Excessive number of operating points (ensures the organization growth and backup).

• Device components (server, switch, and so on). Duplication at the level of critical elements such as:

•Processors (two or more),

• RAM (two or more),

• HDD (two RAID1hard drives + one backup)

• Network cards and adapters (two cards with one or more ports)

• Power supply units (N+N or N+1)

• IT service components. Backup at the device level and the type of their inclusion. For example:

• Two servers in a failover cluster

• Two servers performing the same function (two domain controllers)

• Two servers performing different functions in the normal mode, but can assume the neighboring function in the case of a failure (File server and print server. In case of failure, one can install the function on a neighboring server).

• Geographically separated devices within the site. Allocating a pair of devices in different racks, rooms, or buildings within the same site.

• Geographically separated devices at the site level. The allocation of a pair of devices on different sites.

Determining the backup level depends on the business criticality of failure or downtime. It is analyzed within Risk Management, formalized in documents on equipment standards.

Backup and Archiving Strategy

Backup systems and archiving IT infrastructure may define the following requirements:

• System components must be installed on dedicated physical elements (servers, storage, and so on)

• Backup and archiving systems on the same components can be combined.

• Access rights to (automatic) backup accounts should not provide an opportunity to delete data or overwrite them.

Backup building strategy

Backup component is one of the important elements when building an IT architecture. It relates to the information security, fault tolerance and recovery. The fault tolerance can be as follows:

• Fault tolerance at component redundancy level;

• Fault tolerance at the level of duplicating elements;

• Fault tolerance within the site;

• Fault tolerance with geographically distributed site.