Using Cloud Computing Platforms
Virtualization Technologies and Load Balancing
General Objective (GO):
• Students can able to apply virtualization techniques in cloud
environment with load balancing strategies.
Specific Objectives (SO):
SO1: Identify the network resources in cloud environment to be load
balanced. (U/C)
SO2: To compare and contrast the management with load balancing
and without load balancing strategy. (Ap/C)
SO3: Apply different algorithms in real time cloud environment. (Ap/C)
Virtualization Technologies
• Majority of cloud-based systems combine their resources into
pools that can be assigned on-demand to users.
• Without resource pooling, it is impossible to attain efficient
utilization, provide reasonable costs to users, and proactively
react to demand.
Ajith G.S: poposir.orgfree.com
Virtualization Technologies
• Pooled resources can be accessed using a technique called
virtualization.
• Virtualization assigns a logical name for a physical resource and
then provides a pointer to that physical resource when a
request is made.
• The mapping of virtual resources to physical resources can be
both dynamic and facile; which helps efficient resource
management.
Ajith G.S: poposir.orgfree.com
Virtualization Technologies
• Virtualized components for cloud computing:
• Access: A client can request access to a cloud service from any
location.
• Application: A cloud has multiple application instances and
directs requests to an instance based on conditions.
Ajith G.S: poposir.orgfree.com
Virtualization Technologies
• Virtualized components for cloud computing:
• CPU: Computers can be partitioned into a set of virtual
machines with each machine being assigned a workload.
• Alternatively, systems can be virtualized through load-
balancing technologies.
• Storage: Data is stored across storage devices and often
replicated for redundancy.
• To enable these characteristics, resources must be highly
configurable and flexible.
Ajith G.S: poposir.orgfree.com
Virtualization Technologies
• Mobility patterns for Cloud Computing:
• P2V: Physical to Virtual
• V2V: Virtual to Virtual
• V2P: Virtual to Physical
• P2P: Physical to Physical
• D2C: Datacenter to Cloud
• C2C: Cloud to Cloud
• C2D: Cloud to Datacenter
• D2D: Datacenter to Datacenter
Ajith G.S: poposir.orgfree.com
Virtualization Technologies
• Virtualization is a key enabler for the following attributes of
cloud computing:
• Service-based: A service-based architecture is where clients are
abstracted from service providers through service interfaces.
• Scalable and elastic: Services can be altered to affect capacity
and performance on demand.
• Shared services: Resources are pooled in order to create
greater efficiencies.
• Metered usage: Services are billed on a usage basis.
Ajith G.S: poposir.orgfree.com
Virtualization Technologies Load Balancing and Virtualization
• One characteristic of cloud computing is virtualized network
access to a service.
• No matter where you access the service, you are directed to the
available resources.
• Cloud Load Balancing is the process of distributing workloads and
computing resources in a cloud computing environment.
• It allows enterprises to manage application or workload demands
by allocating resources among multiple computers, networks or
servers.
• Load Balancing. It can be implemented in hardware or in software.
Ajith G.S: poposir.orgfree.com
Virtualization Technologies Load Balancing and Virtualization
• Load balancing is an optimization technique; it can be used to
increase utilization and throughput, lower latency, reduce
response time, and avoid system overload.
• The following network resources can be load balanced:
• Network interfaces and services such as DNS, FTP, and HTTP.
• Connections through intelligent switches.
• Processing through computer system assignment.
• Storage resources.
• Access to application instances.
Ajith G.S: poposir.orgfree.com
Virtualization Technologies Load Balancing and Virtualization
• Load balancing make a system reliable through managed redirection.
• It uses different types of algorithm to decide where the traffic is
routed.
• When a request from a client or service requester arrives, the load
balancer uses a scheduling algorithm to assign where the request is
sent.
• Typical scheduling algorithms in use are Round Robin, Weighted
Round Robin, Fastest Response Time, Least Connections and
weighted least connections
• It also creates a session so that subsequent traffic related to that
session is routed to the same resource.
Ajith G.S: poposir.orgfree.com
Hyper-threading
VM with VM with VM with
one two CPUs one
• Makes a physical CPU appear as CPU CPU
two Logical CPUs (LCPUs)
Enables operating system (OS) to
schedule two or more threads
simultaneously
• Two LCPUs share the same
physical resources
While the current thread is stalled,
CPU can execute another thread
• Hypervisor running on a hyper-
LCPU LCPU
threading-enabled CPU provides LCPU LCPU
improved performance and
utilization Thread 1 and 2 Dual – core Thread 1 and 2
Single – socket system
with hyperthreading
13
CPU Load Balancing
VM with VM with VM with
one one one
CPU CPU CPU
LCPU LCPU
LCPU LCPU
Hyperthreaded dual – core
Dual – socket system
14
Key Network Traffic Management Techniques
1. Balancing client workload: Hardware based
2. Balancing client workload: Software based
3. Storm control
4. NIC teaming
5. Limit and share
6. Traffic shaping
7. Multipathing
Copyright © 2011 EMC Corporation. All Rights Reserved. Virtualized Data Center – Networking 15
Technique 1 – Balancing Client Workload: Hardware
Based
• A device (physical switch/router) distributes client traffic across
multiple servers – physical or virtual machines
• Clients use IP address (virtual) of the load balancing device to
send requests
• Load balancing device decides where to forward request
• Decision making is typically governed by load balancing policy,
for example: Round robin, Weighted round robin, Least
connections
Copyright © 2011 EMC Corporation. All Rights Reserved. Virtualized Data Center – Networking 16
Technique 2 – Balancing Client Workload: Software
Based
• Performed by software running on a physical or virtual machine
Example: DNS server load balancing
Allows multiple IP addresses for a domain name
Maps domain name to different IP addresses in a round robin
fashion
Allows clients accessing a domain name to send requests to different
servers
Example: Microsoft Network Load Balancing
A special driver on each server in a cluster balances clients’ workload
The driver presents a single IP address (virtual) to all clients – each IP
packet to reach each server
The driver maps each request to a particular server – other servers in
the cluster drop the request
Copyright © 2011 EMC Corporation. All Rights Reserved. Virtualized Data Center – Networking 17
Technique 3 – Storm Control
• Prevents impact of storm on regular LAN/VLAN traffic
Storm: Flooding of frames on a LAN/VLAN creating excessive traffic
and degrading network performance
• Counts frames of a specified type over 1-second and compares
with the threshold
• Switch port blocks traffic if threshold is reached and drops the
subsequent frames over the next time interval
Source: “Catalyst 3550
Multilayer Switch
Software
Configuration Guide” –
Cisco Systems, Inc.
Copyright © 2011 EMC Corporation. All Rights Reserved. Virtualized Data Center – Networking 18
Technique 4 – NIC Teaming
• Logically groups physical NICs connected to a virtual switch
Creates NIC teams whose members can be active and
standby
Balances traffic load across active NIC team members
Provides failover in the event of an NIC/link failure
Allows associating policies for load balancing and failover at a
virtual switch or a port group
Copyright © 2011 EMC Corporation. All Rights Reserved. Virtualized Data Center – Networking 19
Technique 5 – Limit and Share
• Are configurable parameters at distributed virtual switch
• Are configured to control different types of network traffic,
competing for a physical NIC or NIC team
• Ensure that business critical applications get required bandwidth
Configurable
Parameter Description
• Sets limit on maximum bandwidth per traffic type
Traffic type will not exceed limit
Limit
• Is specified in Mbps
• Applies to an NIC team
• Specifies relative priority for allocating bandwidth to
different traffic types
Share • Is specified as numbers
• Applies to a physical NIC
Copyright © 2011 EMC Corporation. All Rights Reserved. Virtualized Data Center – Networking 20
Technique 6 – Traffic Shaping
• Controls network bandwidth at virtual/distributed virtual switch or port
group
• Prevents impact on business-critical application traffic by non-critical traffic
flow
Parameter Description
• Data transfer rate allowed over time
Average Bandwidth • Workload at a switch port can intermittently exceed av. Bandwidth
• Burst: When the workload exceeds the average bandwidth, it is called burst
Peak Bandwidth • Max data transfer rate without queuing/dropping frames
• Max amount of data allowed to transfer in a burst
Burst Size • Burst size = bandwidth × time
• Bandwidth in a burst can go up to peak bandwidth
Burst size = bandwidth x time Example:
Peak bandwidth Average bandwidth = 1 Kbps, Peak bandwidth =
Bandwidth
4Kbps, Burst size = 3 Kb
• Burst with data rate 3 Kbps can stay for 1
Average bandwidth second or,
• Burst with data rate 1.5 Kbps can stay for 2
Workload seconds, etc.
Time
Copyright © 2011 EMC Corporation. All Rights Reserved. Virtualized Data Center – Networking 21
Technique 7 – Multipathing
Physical server
Multipathing Hypervisor Kernel
Multipathing
A technique allowing a physical server to use
multiple physical paths for transferring data HBA 1 HBA 2
between the physical server and a LUN on a
storage system.
• Is built into hypervisor or provided by third-
party vendor
• Recognizes alternate I/O path to a LUN and SC 1 SC 2
enables failover
• Performs load balancing by distributing I/O to SC –
Storage
all available paths Controller
LUN 0 LUN 1
Storage array
Copyright © 2011 EMC Corporation. All Rights Reserved. Virtualized Data Center – Networking 22
Virtualization Technologies Advanced Load Balancing
• Workload managers are more sophisticated load balancers.
• They determine the current utilization of the resources in their
pool, the response time, the work queue length, connection
latency and capacity, and other factors in order to assign tasks
to each resource.
• An Application Delivery Controller (ADC) is a combination of
load balancer and application server.
• It is placed between a firewall/router and a data centre
providing the web services.
Ajith G.S: poposir.orgfree.com
Virtualization Technologies Advanced Load Balancing
• ADCs are also referred to as a content switch, multilayer
switch, or web switch.
• An ADC is assigned a virtual IP address (VIP) that it maps to a
pool of servers based on application specific criteria.
• An ADC is considered to be an advanced version of a load
balancer and lowers the workload of the Web servers.
• Services provided by an ADC include data compression,
content caching, security, server health monitoring, Secure
Sockets Layer[SSL] offload and advanced routing based on
current conditions.
Ajith G.S: poposir.orgfree.com
Thank You
