Azure Load Balancing Resources —Simplified use cases with DevOps (CI/CD)

In this blog, the basics of load balancing and we will have a look into the different load balancing feature at different layers of networking for which the Azure provides us different load balancers.

This blog extracts the essence of effective usage of load balancers related to DevOps.

Topics to be covered:

• TCP/IP vs OSI models
• What is a load balancer?
• Load balancing capabilities provided by Azure
• Usage of Traffic Manager in load balancing
• Use-case Architecture to choose the right load balancing option in DevOps

Fig, 1.0. Illustration of load balancing — where balancing time and money !

The communication (send/receive) of data happens in computer using the networking , for that communication to happens the networking relies on the concept of transporting the data from point to point with security features ensured.

Communication with networking can be done using two major models -> TCP/IP or OSI Model, its good to understand the basics of these concepts to choose the load balancing options.

TCP/IP vs OSI model

Fig 1.1 . Layer Differentiation for TCP/IP and OSI models

What is a load balancer ?

Load Balancer are used to disturbed the traffic when its low / heavy accordingly to the backend servers (VM) or Resources.

• It occupies layer 4 (Transport) in OSI model and 3 in TCP/IP model.
• Load Balancers provide security using virtual network and firewalls.

Two major load balancers are available :

1. public -> Internet facing

2. private -> On Prem hybrid connections

Load balancing capabilities provided by Azure

The following table provides a simplified view of the the load balancing option provided by the Azure.

Usage of Traffic Manager in load balancing

When a application is public facing , based on metrics like discounts on sale, popularity and festival the application will be hit with higher frequency and the number of request to the application will be increasing based on the usage which is termed as traffic.

When the Traffic is heavy there are possibilities of

• Delay in loading application
• Improper working of application functionality
• Refresh latencies would affect the money transaction as well.

So ultimately it will lead to the loss in business .

Inorder to avoid this Azure provides us Traffic Manager with capability to reroute the traffic / requests into load balancers with various routing techniques based on

Geographical, latency, weighted, priority, subnet and multi-valued.

Fig1.2. An illustration of a traffic police (traffic manager) managing the Traffic.

Use-case Architecture to choose the right load balancing option in DevOps

1. Load Balancer:

Load Balancer can be used against the Virtual Machine Scale Sets (VMSS) to balance and it directly depends the number of self hosted runner used under deployments.

During Development , multiple teams can run multiple deployments at a time in that case if there is a singe self hosted vm then there runner will be in waiting mode for long which will increase the development teams. so in-order to avoid it based on the necessity its good practice to increase the VMSS for the demand and smoothly run the deployment.

The VMSS can be used in Azure Devops pipelines along with Azure Deployment Groups to facilitates the demand.

Use Case:

Considering there are three different development team which are development different functionality which are with different programming languages and frameworks, they have a deployment on same day to make all the feature to work coherently. In Order to do that the Self hosted Vm should balance the load and provision similar vms .

Fig.2.0: Deployment Architecture for load balancer use case

2. Application Gateway:

Application Gateway is used in the( Layer 7) which used to handle the data at Application layer (Http/Https/Http2).

This will be used to reroute the traffics to application layered resources in Azure like Azure Apim Management / it can provision VM’s for a set users using application gateway for containers.

Use Case:

Create a VMs for the training teams A and B , with two VMs .

Vm1 :

Jenkins training will be given.

When team1 runs the pipeline build and deployment should happen in vm1.

Vm2 :

ADO pipelines training will be given.

When team2 runs the deployment should happen in vm2.

As a DevOps engineer provision VMs for both these teams mentor which will be helpful to train the students.

Fig.2.1: Deployment Architecture for application Gateway use case

3. Azure Front Door:

Azure Front Door uses Point of Presence as a underlying mechanism in provisioning the necessary data in minutes . its a combination of security and Content Delivery networks. Azure Front Door also opens up facilities to connect to onPrem and azure based resource data to be fetched ., so it opens a Hybrid solution.

It can be effectively used in Banking and Media Industries.

Use Case:

There is a global news channel which needs its website to be updated with latest news videos from each region .

The videos are stored in the harmonics account. Now you to fetch the videos into your website securely and deploy them into your web app on daily basis.

Fig.2.2: Fig.2.1: Deployment Architecture for Azure Front Door use case

Conclusion:

I have explained the load balancing importance and the resources which Azure offers to manage load balancing with respect to deployment strategies with DevOps given a simplified examples.

In real world mostly load balancer and application gateway is used widely.

Azure Front has extensible features which focuses on CDN.

Leave a Clap if you liked this Content !!

References:

What is Azure Load Balancer? - Azure Load Balancer

Azure Front Door

Deploy to Azure Kubernetes Service with Azure Pipelines - Azure Kubernetes Service

Quickstart: Direct web traffic using the portal - Azure Application Gateway