Easily Manage grouped of virtual machines using Azure VMSS with same specs and configuration
Azure VMSS - virtual machine scale sets references architecture
Azure virtual machine scale sets, or in short Azure VMSS is a services to provision a group of virtual machine with identical specs and configuration under a pre configured load balancer allowing developer to not worry about their virtual machines scalability configuration.
The virtual machines that was provisioned through VMSS can be configured to automatically scale based two following factors including pre-defined schedule or incoming traffic volume. VMSS support both horizontally (adding new VM with identical hardware specs and confguration) and vertically scale (increase hardware specs including CPU, RAM or storage to all virtual machines on the same group/VMSS).
Apart from scalebility and replicability, azure VMSS can be configured to provision or deploy the VM into separate availability zones or region which will ensure high availability (HA) for the services. When one availability zones or region is down, the load balancer will automatically route incoming requests to the active servers outside the impacted zones/regions.
Key Capabilities of Azure VMSS
Below are summarized capability of Azure VMSS
- Provision a sets or a group virtual machine (up to 1K VM) with identic specs and configuration
- Distributed the VM provisioning location into one or more availability zones
- Automated scale based on volume or schedule
— So.. What kind of use cases suitable for these capabilities of Azure VMSS?
Scaling Agility
A single VM which not provisioned through VMSS has limited automated scaling options. It only capable of automateitcally increase the server resources vertically (RAM, CPU and Storage). In contrast, VMSS can automatically remove or add new VM instances (horizontal scale) depending on the traffic volume or demand.
Such ability will be handy when dealing with dyanmic throughput environment and need to optimize or save the infrastructure cost by automatically minimizing the number of provisioned virtual machine when the traffic is low.
Typical Business Usecases that uses Azure VMSS
Below are the two most common business use cases that definitely can be accelerated by azure VMSS.
Distributed Computing
Distributed computing is the most typical use cases that can greatly benefited with azure virtual machine scale sets (VMSS) capabilities becasue distributed computing will require to utilize a sets of virtual machine to complete the business transaction/logic as the transaction will be broadcasted or distributed into different virtual machine to achieve high reliability, high throughput and high reliability.
Most common distributed computing use is a high throughput backend REST API server that serve B2C frontend/customer facing channel (can be web app or mobile app) on various industry (e-commerce, banking, telco, etc).
for a back offices application that serves B2B (business to business) use cases such as order management, production planning and so on are generally not get much benefit from Azure VMSS as the number of request throughput are usually low (unless there is an requirement to make it high availability, if HA is a requirement azure VMSS is the best in slot solution)
Machine Learning Training
VM provisioned through Azure VMSS can be equipped with GPU (Graphical processing unit) such as like Nvidia A100 & RTX 5000 series to allow data scientist to train their model (Assuming the training will be GPU intensive instead of GPU intensive).
— Typically these ML will solve engineering or process simulation complex problem or even LLM (large language model training).
