30 lines
1.2 KiB
Markdown
30 lines
1.2 KiB
Markdown
---
|
|
layout: default
|
|
title: Deployment - Kubernetes
|
|
parent: Deployment
|
|
nav_order: 3
|
|
---
|
|
|
|
# Deployment on Kubernetes
|
|
|
|
<p>
|
|
<img src="../images/logos/kubernetes.png" width="100" style="padding-right: 10px">
|
|
</p>
|
|
|
|
## UNDER CONSTRUCTION
|
|
|
|
The following areas are actively being worked on for Kubernetes:
|
|
1. Detailed Documentaiton
|
|
2. [Helm Chart (v3)](https://github.com/authelia/chartrepo)
|
|
3. Kustomize Deployment
|
|
4. Manifest Examples
|
|
|
|
Users are welcome to reach out directly on our [Matrix Room](https://riot.im/app/#/room/#authelia:matrix.org) or
|
|
[Discord Server](https://discord.authelia.com) if they are looking for help setting up on Kubernetes in the meantime.
|
|
|
|
## FAQ
|
|
|
|
### RAM usage
|
|
|
|
If using file-based authentication, the argon2id provider will by default use 1GB of RAM for password generation. This means you should allow for at least this amount in your deployment/daemonset spec and have this much available on your node, alternatively you can [tweak the providers settings](https://www.authelia.com/docs/configuration/authentication/file.html#memory). Otherwise, your Authelia may OOM during login. See [here](https://github.com/authelia/authelia/issues/1234#issuecomment-663910799) for more info.
|