CS2232- IT Infrastructure

Management and Spreadsheet-

Based Data Analysis
Module - II
Data Center Systems and
Server
Agenda
 Datacenter introduction and history
 Datacenter buliding blocks
 Datacenter availability
 Datacenter performance
 Datacenter security
 Servers introduction and history
 Server building blocks
 Server availability
 Server performance and security

CS1003 - Programming in C
DATA CENTERS

Introduction
• A centralized location where computer related resources (and data) are
stored.
• A datacenter provides power supply ,cooling, fire prevention and detection
equipments racks, and other facilities for the installed infrastructure
components.

3 / 39
Introduction-Location of the Datacenter

• Many variables should be considered to determine the where a datacenter

could be installed.
Installation Category:-
Environment
Visibility
Utilities

4 / 39
 Environment
• Is the enough space available to expand the datacenter?
• Is the location vulnerable to flooding?
• Is the datacenter located in a cyclone prone area?
• What is the chance of an earthquake?
• What is the climate like?

5 / 39
 Visibility
• Is the location of the datacenter included in public map? (like
[Link]
• Does the building have windows?
• Are marking on the building showing that this building contains datacenter?

6 / 39
 Utilities
• Is it possible to have independent power provider and internet provider?
• Can cabling routes to the building be determined?
• Is the datacenter located in a shared building?
• Is the enough power available to supply the datacenter?
• Is cheap power supply available?
• What is available bandwidth of the external data connection?

7 / 39
Introduction-Physical structure of Datacenter
• The Physical structure of
Datacenter contains floors, wall,
windows, doors, water and gas
pipes.
• Dig:Layout of the datacenter

8 / 39
 Disadvantage
• Raised floors are expensive.
• The total available height in the datacenter is decreased.
• The maximum floor load is limited.

9 / 39
Types of Data Centers

Data Center

Enterprise Data Colocation Managed Cloud Edge

Centers

(owned) (Shared) (outsource) (virtual) (Near User)

CS1003 - Programming in C
 Building Blocks

These building blocks can be grouped into five major categories:

1. IT Infrastructure
2. Network Infrastructure
3. Power Systems
4. Cooling & Environmental Control
5. Physical Security & Monitoring

CS1003 - Programming in C
 Building Blocks

Building Block Purpose Examples

IT Infrastructure Data processing & storage Servers, storage, virtualization

Network Infrastructure Connectivity & traffic management Switches, routers, firewalls

Power Systems Continuous power supply UPS, PDUs, diesel generators

Cooling Systems Maintain temperature & humidity CRAC, chillers, immersion cooling

Security Systems Protect from unauthorized access CCTV, biometrics, fire suppression

CS1003 - Programming in C
Building blocks – Datacenters Availability NFR’s

Availability Tier Description Availability Key Features

• Tiering (Uptime institute) Tier 1 Basic capacity 99.671%

Single path for power/cooling, no
redundancy
• Only for the datacenter, not the IT
• Almost no Tier 4 datacenters in Tier 2
Redundant
99.741%
Redundant power/cooling
components components, single distribution path
use

Multiple independent power/cooling

Concurrently
Tier 3 99.982% paths, maintenance without
maintainable
downtime

Fully redundant, multiple

Tier 4 Fault-tolerant 99.995% independent paths, can sustain
failures without affecting operations
 Datacenters Performance
Datacenter performance refers to the efficiency, speed, and reliability with which a datacenter
processes, stores, and transmits data.
Key factors influencing performance include hardware, software, and network design.
1. Compute Performance:-
• CPU & Processors: Number of cores, clock speed, and architecture (e.g., x86, ARM).
• Memory: RAM size, type (DDR4/DDR5), ECC support, and bandwidth.
• Virtualization: Efficient hypervisors improve utilization and performance of multiple workloads.
2. Storage Performance
• Type: SSDs are faster than HDDs. NVMe drives offer high-speed data access.
• Redundancy: RAID configurations improve reliability but may impact write performance.
• IOPS & Latency: Higher IOPS and lower latency improve application responsiveness.
3. Network Performance
• Bandwidth: Network interface speeds (1G, 10G, 25G, or higher).
• Redundancy: Multiple network paths reduce downtime.
 Datacenters Performance
• Switching & Routing Efficiency: Low-latency switches enhance throughput.
4. Cooling & Power Efficiency
• Thermal Management: Proper cooling ensures hardware runs at optimal speed.
• Power Supply Redundancy: Prevents performance degradation during power issues.
5. Monitoring & Optimization
• Performance Metrics: CPU utilization, memory usage, storage I/O, network throughput.
• Load Balancing: Distributes workloads evenly across servers.
• Caching & Acceleration: Improves response time for frequent requests.
6. Reliability & Availability
• Redundancy: Failover mechanisms ensure continuous operation.
• Maintenance: Scheduled updates without impacting uptime.
Datacenter security
Physical security
• Physical access control
• Cameras
• Security guards, dogs

Datacenter secure is mostly a matter of physical security ,ensuring that

equipment is physically safe behind the datacenter doors.
To implement this, doors, windows, and other entry point
must be secured.

16 / 39
 Servers introduction and history
• The first general purpose universal computer was the ENIAC.
• The Electronic Numerical Integrator And Computer (ENIAC) was designed in 1943 and was financed by the
United States Army in the midst of World War II.
• The machine was finished and in full operation in 1946 (after the war) and was in continuous operation until
1955.
• First computing machines were actually mechanical calculators. However, those were not real computers.
Real computers have two properties:
 They calculate and they are programmable.
 Making programmable computers became only useable in practice after the invention of punched cards.
and Using punched cards the computers could be programmed and reprogrammed easily.
Cont..ENIAC
The ENIAC was very large in size and complexity.
It contained:-17,468 vacuum tubes,7,200 crystal diodes
1,500 relays,70,000 resistors
10,000 capacitors,Around 5 million hand-soldered joints.
The ENIAC weighed about 30 tons ,and consumed 150 kW of
power.
• The ENIAC got its input using an IBM card reader, and IBM
punched cards were used for output as well.
• In the 1960s computers started to be built using transistors
instead of the large and heavy vacuum tubes used in earlier
computers.
• Transistor- based machines were smaller, faster, cheaper to
produce, required less power, and were more reliable.
• The transistor-based machines were followed in the 1970s by
machines based on integrated circuit (IC) technology. They
provided standardized building blocks like AND gates, OR
gates, counters, adders, flip flops, etc.
Server building blocks
It Includes
• Processor
• x86 processor architecture
• 64 bit processor
• Memory Interface
• External interface
• Internal interface
• x86 platform
Processor
Role of the CPU (Central Processing Unit):-Acts as the brain of the computer/server.
Executes instructions from programs and operating systems.
Performs fetch → decode → execute → write-back cycle.
Handles arithmetic/logic operations, control signals, and data movement.
Clock Speed: Measured in Hertz (Hz), usually GHz today.
Determines how many instruction cycles the CPU can perform per second.
Higher clock speed ⇒ faster instruction processing (but also depends on architecture and cores).
Cores:
A core is an independent processing unit inside the CPU.
Single-core: Can execute one instruction thread at a [Link]-core (dual, quad, hexa, octa, etc.):
Each core can handle its own thread(s).
Enables parallel processing for multitasking and server [Link] servers use multi-core, multi-
threaded CPUs for scalability.
 x86 processor
Origin : Intel 8086 family; standard for PCs and data-center servers.
Type : CISC (Complex Instruction Set Computing) – rich, versatile instruction set.
Compatibility: Backward-compatible with earlier x86 CPUs.
Modes : Real (16-bit), Protected (32-bit), Long (64-bit).
Registers : General-purpose (AX, BX, etc.), segment registers, instruction pointer.
Virtualization: Intel VT-x / AMD-V support for multiple VMs.
Usage : Widely adopted (Intel Xeon, AMD EPYC) for cloud, enterprise, and HPC servers.
 processor
64 bit processor
• Handles 64-bit wide data and memory addresses.
• Supports large memory (theoretically up to 16 EB).
• Required by modern operating systems and large-scale applications.
• Examples: Intel Xeon, AMD EPYC.

CS1003 - Programming in C
 Memory Interface
• Connects CPU ↔ RAM.
• Types: DDR4/DDR5, ECC (Error-Correcting Code) for reliability.
• Key factors: bandwidth (GB/s) and latency for faster data access..
External interface
• Provide connectivity to outside devices/networks.
• Includes network ports (Ethernet 1/10/40/100 GbE), USB, PCIe slots, HDMI/Display (where
needed).
Internal interface
• High-speed paths inside the server.
• System bus, memory bus, storage buses (SATA, NVMe), PCIe backplane.
• Enable fast CPU–RAM–storage communication.
Server-x86 building blocks
• Many modern server motherboards integrate
Northbridge functions into the CPU.
• so the traditional Northbridge–Southbridge
separation is less visible.
• Interfaces like PCIe, SATA, USB still exist, but ISA is
obsolete.
• RAM and storage interfaces (NVMe, DDR4/DDR5) are
critical.
• Network interfaces (NICs) are not shown but are
essential for servers.
 Server-performance
Definition: Moore’s Law observes that the number of transistors on a microprocessor doubles
approximately every 18–24 months, leading to exponential growth in computing power over
time.
Implication: More transistors → faster CPUs, more cores, larger caches → better performance.
Importance for Server Performance:
Processing Power
• More transistors allow faster processing and support multi-core architectures.
• Servers can handle more simultaneous tasks, improving throughput.
 Cont..Server-performance
Energy Efficiency
• Advanced transistor technology can reduce power consumption per computation.
• Helps in data centers where energy cost is a major concern.
Memory and Cache
• Moore’s Law enables larger and faster caches.
• Reduces bottlenecks in data access for server applications.
Virtualization and Cloud
• Servers benefit from increased cores and higher CPU performance to host more virtual
machines efficiently.
• Supports scalable cloud workloads.

CS1003 - Programming in C
 Server Availability
• Server availability refers to the ability of a server to remain operational and accessible to users when needed. It
is a key Non-Functional Requirement (NFR) in datacenter design and IT services.
• 1. Key Metrics
• Uptime: Percentage of time the server is operational (e.g., 99.9%, 99.99%).
• Mean Time Between Failures (MTBF): Average time between hardware failures.
• Mean Time to Repair (MTTR): Average time to restore service after a failure.
2. Techniques to Improve Availability
Redundancy:
• Duplicate critical components (power supplies, network interfaces, storage).
• Prevents single points of failure.
Failover Systems:
• Automatic switching to standby servers during failure.
 Server Availability
Load Balancing:
Distributes workload to prevent overloading a single server.
Clustering:
Multiple servers work together; if one fails, others take over.
Regular Backups & Recovery Plans:
Ensures quick restoration in case of data loss.
3. Monitoring
 Continuous monitoring of hardware, software, and network health.
 Alerts for potential failures to enable proactive maintenance.

CS1003 - Programming in C
 Server Performance
Server performance refers to how efficiently a server executes workloads, handles requests,
and delivers services to clients.
It is influenced by hardware, software, and network factors.
1. Compute Performance
CPU: Number of cores, clock speed, and architecture (x86, ARM).
Memory: Size, type (DDR4/DDR5), bandwidth, and ECC support.
Multi-threading & Virtualization: Efficient handling of multiple tasks and VMs.
2. Storage Performance
Disk Type: SSDs > HDDs; NVMe drives offer higher speed.
IOPS & Latency: Determines speed of data access.
RAID & Caching: Balances performance and reliability.

CS1003 - Programming in C
 Server Performance
[Link] Performance
Bandwidth & Speed: High-speed NICs (10G, 25G, 40G) improve data transfer.
Redundancy: Multiple network paths prevent downtime.
Latency: Low-latency switches and optimized routing enhance responsiveness.
4. Software & OS Optimization
Efficient operating system and server software improve resource utilization.
Load balancing and caching improve response times.
5. Monitoring & Maintenance
Continuous monitoring of CPU, memory, storage, and network.
Regular updates and proactive maintenance prevent performance degradation.

CS1003 - Programming in C
 Server Security

• Disable External USB Ports in BIOS – Prevents unauthorized use of removable media.
• BIOS Password Protection – Secures BIOS settings from unauthorized changes.
• Chassis Intrusion Detection – Some servers can detect and log when the server
housing is physically opened.
Thank You