Container groups

Container groups contain a set of containers that share their life cycle, resources, network configuration, and storage resources, essentially grouping them together. Containers within a container group are scheduled together on the same host (the server on which these run) and in ACI, which is the top-level resource. Container groups can consist of one or more containers. The following diagram is an example container group that contains two containers scheduled on the same host. The container group exposes port 80 on a single public IP (an IP accessible from the internet) and has a public DNS label associated with it. The container group also contains two different storage accounts, each of which is associated with one of the containers. Of the two containers, one container exposes port 80 while the other exposes port 1433. On private ports, as illustrated in the following diagram, private refers to ports, IP, or DNS being restricted to access within the local network only:

Figure 11.2 – Example container group

You now have a basic understanding of what a container group is, so next, we will explore some common scenarios for using multi-container groups.

Common scenarios

There are several scenarios where multi-container groups prove beneficial, such as when you define a frontend and backend for your container groups. Tasks may be divided into an application container and a logging or monitoring container; they perform two separate functions but are both complementary to the end service delivery of the application. Images can be deployed and managed by different teams within your organization.

Some other scenarios include the following:

• An application container and a database container. The application container(s) may deliver the end application, while the database container commits data to a service location.

• A web application service deployed to a container within the group and another container for managing source control.

Resources

Container groups are allocated resources by ACI by adding resource requests for all instances in a container group, such as CPUs, memory, and even graphics processing units (GPUs). For example, if you have two container instances in a container group and each instance requires two CPUs and 4 gigabytes (GB) of memory, the container group will then be allocated four CPUs and 8 GB of memory, as depicted in the following diagram:

Figure 11.3 – Container group resource allocation

You now understand how resources are allocated for container groups. Next, we will explore the networking components of the Azure Container Instance service.