Hi, everyone, and welcome to this lecture. In this lecture, I'm going to talk about instance groups. An instance group is actually a collection of VM instances. The key advantage of an instance group, however, is the ability to manage all of the instances as a single entity. Google Compute Engine offers two kinds of VM instance groups. They include managed instance groups, or MIGs, and unmanaged instance groups.

Using a managed instance group allows you to run applications on several identically configured virtual machines. Additionally, a managed instance group offers several automated services such as autoscaling, auto-healing, regional deployment, and auto updating. By leveraging the many automated services that are offered by a managed instance group, you can make workload scalable and highly available.

Typically you'd want to use a managed instance group for stateless workloads, like maybe a website frontend. Managed instance groups are also a good choice for high-performance workloads and batch workloads, such as image processing from a queue. It's important to note that each VM instance within a managed instance group is created from an instance template. If you prefer - or need - to load balance apps across a fleet of virtual machines that you need to manage yourself, you might want to consider unmanaged instance groups.

An unmanaged instance group contains multiple heterogeneous VM instances. You can, as needed, add and remove instances from the group. As you would expect, an unmanaged instance group does not offer any of the features that a managed instance group provides. You get no autoscaling, no auto-healing, and no rolling update support. As a rule of thumb, you should only consider using unmanaged instance groups if you need to apply load balancing to groups of heterogeneous instances, or if you need to manage the instances within the group yourself.