This page describes the tools and requirements needed to deploy Polyaxon, to provide some more contextual information.
In order to run Polyaxon you need some computational resources,
- Disk space
- Networking (both internal and external)
- Creating, resizing, and deleting clusters
these resources could be your personal computer, university server, or some other organization that hosts computational resources that can be accessed remotely.
Polyaxon will work fine with a cloud provider or a custom cluster deployments as well. For these materials, any cluster with Kubernetes installed will work with Polyaxon.
Container technology is essentially the idea of bundling all of the necessary components to run a piece of software. There are many ways to do this, in the case of Polyaxon we use Docker to bundle the user's dependencies required to run the code.
Kubernetes is a service that runs on cloud infrastructures. It provides a single point of contact with the machinery of your cluster deployment, and allows a user to specify the computational requirements that they need (e.g., how many machines, how many CPUs per machine, how much RAM). Then, it handles the resources on the cluster and ensures that these resources are always available. If something goes down, kubernetes will try to automatically bring it back up.
Helm is the official package manager for Kubernetes, and a way of specifying kubernetes objects with a standard template.
Polyaxon is installed into Kubernetes using Helm.
Polyaxon is an open-source system that abstracts and simplifies training, monitoring, and scaling deep learning applications on Kubernetes clusters.
In order to run Polyaxon on top of Kubernetes, you need:
- Access to internet to pull images, unless you have all the images you need for running your experiments accessible to Kubernetes through a locally deployed registry.
- Kubernetes with a version >= 1.8.0.
- Helm with a version ≥ 2.5.
- Persistence for data, outputs, logs, and code.
!!! note Although Polyaxon might work with previous versions than those recommended by us, we officially don't support them, and we don't recommend using them.