Microsoft Azure Availability Sets

In this blog, we will explain the basic concept of Microsoft Azure Availability Sets.

THINGS TO KEEP IN MIND REGARDING AVAILABILITY SETS

Group the related VMs into availability sets especially for multi-tier application.
Combin a load balancer with availability sets to make sure that the user does not face any interruptions.
In case of using traditional storage account, make sure that you are using separate storage accounts for each availability set.
Consider scale sets for bigger compute jobs.
Availability Set and Scale Set incur runtime charges.

To understand the need for the availability sets, we need to consider the two types of Azure maintenance events. They are Planned and Un-Planned. These events will happen in Microsoft Azure datacenters.
Planned Maintenance – The Azure team gives you the advance notifications about patching, hardware replacements, and other known events.
Un-planned Maintenance – It’s like rack level (or) datacenter level failures which Microsoft did not anticipate.

Fault Domains (FD) are essential for rack servers. There will be VMs that share the same power source and switch. For example, if you have 2 VMs in an availability set, Azure will provision them in two different racks. So if a network (or) power failure, one rack will remain unaffected and your application will be online.
Update Domains (UD) are routing patching and maintenance of the physical host where your VM resides. Also, the physical host needs to reboot to complete the patching and maintenance. During this time azure will migrate your VMs to another host with any service interruption.
The best practice is to keep the VMs of each application tier into their own availability sets.

Thanks for reading this blog. We hope it was useful for you to learn basic concepts about Azure availability sets.

Loges