Scribd
Upload
5 views

ml Assignment-2

Uploaded by

SUJITHA M
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
5 views

ml Assignment-2

Uploaded by

SUJITHA M
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 3

Here’s a simple explanation comparing physical clusters and virtual clusters in cloud

computing:

Physical Clusters

 What are they? A group of physical machines (servers) connected together to work
as a single unit.
 Where are they? Located in data centers (on-premises or in the cloud).
 Performance: Dedicated hardware, so performance is predictable and high.
 Flexibility: Less flexible—scaling up requires adding more physical machines, which
takes time and resources.
 Cost: Higher upfront costs for hardware and maintenance.
 Use Cases: Often used for high-performance computing, scientific simulations, or
workloads requiring guaranteed resources.

Virtual Clusters

 What are they? A group of virtual machines (VMs) or containers that run on shared
physical hardware.
 Where are they? Managed on cloud platforms or on virtualized on-premises
infrastructure.
 Performance: Shares resources with other virtual systems, so performance can vary
slightly based on the load.
 Flexibility: Highly flexible—can quickly add or remove VMs to scale up or down.
 Cost: Pay-as-you-go model, lower upfront cost.
 Use Cases: Commonly used for web applications, development/testing environments,
and workloads needing scalability.

Key Differences in Plain Terms

1. Physical = Real Machines


Virtual = Pretend Machines (built on top of real ones).
2. Scaling:
Physical clusters are like adding new desks to a small office (slow).
Virtual clusters are like rearranging shared desks in a coworking space (quick and
flexible).
3. Cost:
Physical clusters = Buying furniture upfront.
Virtual clusters = Renting furniture monthly.
Virtual Clusters and Resource Management

1. Definition:
o Virtual clusters are like traditional physical server clusters but use Virtual
Machines (VMs) instead. They are built across two or more physical servers.

2. Features of Virtual Clusters:


o Interconnected VMs: VMs in a virtual cluster can communicate using a
virtual network interface card (NIC), enabling seamless collaboration.
o Dynamic Provisioning: VMs can be dynamically added or removed from a
virtual cluster based on resource requirements.
o Fault Tolerance: VMs can replicate across multiple servers to ensure high
availability, disaster recovery, and parallelism.
o Host Migration: VMs can move from one host to another (live migration) for
load balancing or maintenance.
o OS Flexibility: Virtual cluster nodes can run different operating systems,
making them versatile.
3. Management of Virtual Clusters:
o Guest-Based Manager: The cluster manager resides inside a guest OS.
Example: Linux clusters running on Xen hypervisors.
o Host-Based Manager: The cluster manager resides on the host system.
Example: VMware High Availability (HA) system.
o Independent Manager: A standalone manager that works with both guest and
host systems but adds complexity.
o Integrated Manager: A unified manager that handles both physical and
virtual resources.

Disaster Recovery with Virtualization

1. Purpose:
oVirtualization ensures critical business data is safe from disasters like
hardware failures, cyberattacks, or natural calamities by maintaining backup
systems and data.
2. Advantages:
o Quick Recovery: Image-based backups enable faster restoration of systems,
reducing downtime.
o Off-Site Backups: Ensures at least one copy of data is stored off-site for quick
access during emergencies.
o Data Duplication: Businesses can duplicate data in real time for remote
access or recovery.
o Testing Disaster Plans: Virtualization allows businesses to create test
environments to ensure their disaster recovery plans are effective.
o Fault Tolerance: If hardware fails, virtualization maintains functionality
through backup systems.

Comparison of Physical and Virtual Clusters:

Aspect Physical Clusters Virtual Clusters


Node Type Only physical machines Both physical and virtual machines
OS Compatibility Same OS across all nodes Different OSes on various nodes
Failure Impact Node failure affects the cluster VM failure doesn't harm the host
Scalability Limited by hardware Can dynamically grow or shrink
Management Flexibility Requires manual intervention Automated with virtualization tools

You might also like