IT Infrastructure

Architecture
Infrastructure Building Blocks
and Concepts

Infrastructure Lifecycle
Agenda

● Introduction
● Purchasing infrastructure and services
● Deploying the infrastructure
● Testing the Infrastructure
● Maintaining the infrastructure
● Monitoring
● Logging
2
Introduction
● The infrastructure lifecycle
encompasses the following
steps:
 Purchasing infrastructure and
services
 Build-up, testing and go-live
 Maintaining the infrastructure
 Deploying applications
 Decommissioning the
infrastructure
3
Purchasing Infrastructure and Services
● Most large-scale IT projects require procurement of
hardware, software, or services
● The purchase process entails:
 Determining what is needed – BoM (Bill of Material) , SoW (
Statement of Work)
 Getting an offer – Suppliers choice, Bidding and Tendering
 Ordering – Purchasing Department
 Delivery – Checking at delivery time
 Warranty – Period of warranty claims
 Renewal – Goods need renewal after certain time
4
Deploying the Infrastructure
● Assembling the Infrastructure: Infrastructure can be assembled using the following checklist:
 Build up the physical datacenter room
 Install redundant power cabling
 Install racks
 Test the facilities
 Install the server, networking, and storage hardware
 Allow for a burn-in period to ensure the equipment is not “dead on arrival” (DOA) or fails
within the first day
 Check the power and cooling usage of the equipment
 Configure the infrastructure components
 Install systems management tools
 Test systems management processes.
 Provide as-built documentation to the systems managers.
5
Testing the Infrastructure
● After assembling the infrastructure, it should be tested
● Each test type has a predefined scope:
 Functional tests
 Performance tests
 Security tests
● Test Stages
 Unit
 System integration test
 Fallback test
 Migration test
 Acceptance test
6
Go Live Scenarios
● Big Bang
● Parallel Change Over
● Phased Changeover
● At the go-live date, high alert is needed from:
 The project team
 Systems managers
 Service desk
 Senior management

● After the new system is live, on-site support should be available for some
predefined time
7
Maintaining the Infrastructure

● Systems Management Processes: A typical infrastructure

project takes a couple of months to complete. The
infrastructure is often used in operation for many years,
sometimes even decades
● Examples of systems management processes:
 Incident management
 Change management
 Configuration management
● Examples: COBIT and ITIL
8
DevOps for Infrastructure
● DevOps is a contraction of "developer" and "system operator“
● DevOps is typically used by teams developing and running
functional software
● DevOps teams consist of:
 Developers
 Testers
 Application systems managers
● Each team is responsible for developing and running one or
more business applications or services
9
Monitoring
● Monitoring continuously inspects IT components for events
● Examples of events:
 Error conditions
 Signs of (upcoming) failures
 A disk with only little free space left
 Unusually high CPU utilization
 Extreme network bandwidth usage

● Alarms can be sent if a certain threshold is reached

● Monitoring systems provide real-time dashboards with overviews of an entire
infrastructure landscape
● Examples: Nagios, Zabbix, HP Operations Manager, BMC Patrol
10
11
Logging
● Most infrastructure components generate log data
● Examples:
 Network routers and switches
 Operating systems
 Applications
 Databases
 Firewalls
 Intrusion detection systems

● Log data can be used to:

 Correlate events
 Identify sources of application issues
 Identify trends to predict or even prevent unavailability
 Find security vulnerabilities or security breaches

12
Logging
● Logging can generate large amounts of data every day
● Logging detail is usually configurable
● Timestamps of log entries must match exactly to be able to correlate logs from various
sources
● Big data reporting tools can be used to create overviews and to find anomalies
● Analyzing log files is something fundamentally different than monitoring
 Monitoring systems are real-time systems
 Log files are meant for analyzing situations afterwards
● Log files must be properly secured
 Confidentiality: They contain much sensitive data
 Integrity: Using them as evidence
13
Capacity Management
● Capacity Management ensures the timely availability of sufficient
infrastructural capacity to process, transport, and store data now
and in the future
● The following input is needed:
 Monitoring of resources to detect trends
 Reduced free disk capacity provides insight in when to purchase or free-up disk capacity
 Business plans to anticipate on business changes that might have impact on
the infrastructure
 A marketing campaign during the summer time could justify temporary adding server capacity
 Developments in technology
 Upgrading servers when a higher capacity server blade becomes available

14
Deploying Applications
● DTAP:
 Development
 New software is developed or existing software is modified
 Test
 Software is tested by independent testers
 Acceptance
 Software is accepted by a delegation of the user population
 Production
 When all tests are successful, the software is deployed in the production environment

DTAP environments are used in the software development process

DTAP suggests four environments, but often more environments are used

15
Deploying Applications
● A real-world DTAP environment could contain a:
 Sandbox environment – Pre-development environment, where preliminary tests can be performed on
new technology or solutions
 Development environment – To develop new software and configurations
 Test environment – Functionally test new software releases
 User Acceptance test environment – Allow end users to functionally test new releases
 Non-Functional acceptance environment – Setup to be identical to the production environment to
enable reliable performance, availability, and security testing
 Hot Fix environment – Find fixes for production problems, Test fixes before they are deployed to
production
 Production environment – Runs the actual software for end users
 A Systems Management environment is often used to manage the other
environments

16
Decommissioning

● At the end of its lifecycle, infrastructure must

be decommissioned
● The decommissioning process can be broken
down into:
 Preparation
 Execution
 Clean-up
17