Overview

Endpoints are container instances running within Endpoints AI. These instances can be sourced from a container registry such as Docker Hub, GitHub Container Registry, Amazon Elastic Container Registry, or another compatible registry.

Note: When constructing an image for Endpoints AI, utilize the flag --platform linux/amd64,linux/arm64 to ensure compatibility with the platform.

Understanding Endpoint Components and Configuration

An Endpoint is a server container that you create to access hardware resources, assigned with a dynamically generated identifier. For instance, an identifier like 2s56cp0pof1rmt denotes a specific instance.

An Endpoint comprises several components:

1. Container Volume: Contains the operating system and temporary storage. This storage is volatile and will be lost if the Endpoint is halted or rebooted.

2. Disk Volume: Provides permanent storage preserved for the duration of the Endpoint's lease, akin to a hard disk. This storage is persistent and remains available even if the Endpoint is halted or rebooted.

3. Network Storage: Functions similarly to a volume but can be relocated between machines. When utilizing network storage, deleting the Endpoint is the only action available.

4. Ubuntu Linux Container: Capable of executing nearly any software that can run on Ubuntu.

5. Allocated vCPU and System RAM: Resources dedicated to the container and any processes it runs.

6. Optional GPUs or CPUs: Tailored for specific workloads like CUDA or AI/ML tasks, though not mandatory for container initiation.

7. Pre-configured Template: Automates software installation and settings upon Endpoint creation, facilitating one-click access to various packages.

8. Proxy Connection for Web Access: Allows connectivity to any open port on the container. For instance, the format for accessing an Endpoint would be https://[endpoint-id]-[port number].proxy.endpointsai.net.

To begin, refer to the instructions on choosing an Endpoint and then proceed to the guidelines for managing Endpoints.