INTRODUCTION
What is Load Balancing
Load balancing refers to efficiently distributing incoming network traffic across a group
of backend servers, also known as a server farm or server pool.
Modern high-traffic websites must serve hundreds of thousands, if not millions, of concurrent
requests from users or clients and return the correct text, images, video, or application data, all in
a fast and reliable manner. To cost-effectively scale to meet these high volumes, modern
computing best practice generally requires adding more servers.
A load balancer acts as the “traffic cop” sitting in front of your servers and routing client
requests across all servers capable of fulfilling those requests in a manner that maximizes speed
and capacity utilization and ensures that no one server is overworked, which could degrade
performance. If a single server goes down, the load balancer redirects traffic to the remaining
online servers. When a new server is added to the server group, the load balancer automatically
starts to send requests to it.
In this manner, a load balancer performs the following functions:
Distributes client requests or network load efficiently across multiple servers
Ensures high availability and reliability by sending requests only to servers that are
online
Provides the flexibility to add or subtract servers as demand dictates
1
Figure 1 :Load balancer diagram
Load Balancing Types
The load balancer can be hardware based, software based or cloud-based service.
Hardware based solutions comes up as machines with specialized processors, loaded with
proprietary software. To cope up with increasing traffic, you have to buy additional resources
(machines) from the vendor. Software solutions are less expensive and more flexible. We can
install the software on the hardware of our choice or in cloud environments like AWS EC2. The
cloud-based solutions are the managed services available in cloud providers like AWS, Google
Cloud, and Azure etc. These services follow a pay per usage model, which makes them the low
cost in all these three.
How load balancers work
1. A client, such as an application or browser, receives a request and tries to connect with a
server.
2. A load balancer receives the request,and, based on the preset patterns of the algorithm, it
routes the request to one of the servers in a server group (or farm).
3. The server receives the connection request and responds to the client via the load
balancer.
4. The load balancer receives the response and matches the IP of the client with that of the
selected server. It then forwards the packet with the response.
5. Where applicable, the load balancer handles SSL offload, which is the process of
decrypting data using the Security Socket Layer encryption protocol, so that servers don’t
have to do it.
6. The process repeats until the session is over.
2
Figure 2 : Load balancers detect the health of back end resources and do not send traffic to
servers that are not able to fulfill request.
Suppose if the load balancer itself fails, what happens? In this scenario a single instance
of the load balancing server would cause the entire system outage. So here come the distributed
load balancers. We can have multiple load balancers that manages the traffic. There are many
ways to achieve this.
1. Active/Passive load balancers - One load balancer handles the traffic for a site, if that
goes down another passive load balancer will take charge and handles the incoming requests. But
this set up cannot be a good solution for sites with large traffic.
2. Active/Active load balancers - The traffic will be configured in many servers. An
algorithm will choose the load balancer to which the traffic needs to be sent. Other nodes will be
discarded.
3
Load Balancing Process
When load balancing is active, traffic becomes more evenly distributed to monitoring tools. This
ensures the monitoring tools are not overwhelmed, and network visibility can be improved.
1. Load balancing has different meanings in networking. The definition of load
balancing can differ depending on the network device. In the Matrix, load balancing
refers to forwarding network port traffic as evenly as possible to tool ports. The network
traffic, measured by volume of network conversations or packets, is distributed as evenly
as possible to tool ports connected to the rule.
2. Use load balancing when you have limited or no visibility to a faster network
interface. Consider this scenario: your organization has a 10 Gb network link, but your
analysis tools only have 1 Gb interfaces. Until the analysis tools are upgraded, you have
limited visibility into the network. One solution can be to use load balancing. For
example, a 10 Gb network link that is consistently utilized at 30% or less could be load
balanced to three 1 Gb tool ports. Load balancing can play an important role in giving
your organization the network visibility that a restrictive budget or other factor prohibits.
3. Always try dedicating enough tool ports to balance the traffic without dropping
packets. Try dedicating a sufficient number of tool ports when load balancing; the exact
number depends on many factors. You risk dropped packets (tool port oversubscription)
without enough tool ports in a load balancing setup. Load balancing can still be used with
an insufficient number of tool ports, but you might need to enforce packet
trimming or filtering to lower the utilization enough so that packets do not drop.
4. Load balancing does not provide any type of redundancy or failover for your
connected tools. Although you could design a layout to replicate traffic and forward it to
multiple tool ports, a vital step in creating a simple redundancy strategy, load balancing is
not designed to complement this goal. Replication (meaning load balancing is disabled) is
4
a better choice when many tools need identical data. Load balancing guarantees that the
load-balanced traffic forwarded to tools is never identical streams.
5. Load balancing does not interact with applications to achieve results. The purpose of
load balancing is for taking traffic and distributing it more evenly to the analysis tools
connected to tool ports. The Matrix is designed to perform load balancing without agent
software or other potential points of failure. With the exception of packet trailers and
recalculated CRC values after trimming (both disabled by default), the Matrix does not
modify packets.
Figure 3 : Load Balancing Process
5
CONCLUSION
In a large-scale enterprise application, we need to route the traffic to all server nodes.
These requests should be forwarded such that the optimal system performance should be
achieved. As we seen in this tutorial, the load balancers come to our rescue.
They work on different methods or algorithms that routes the traffic to down level
servers. The load balancers can be applied to web servers, database servers, etc. We may have to
go for a set of load balancers in order to avoid a single point of failure.