Skip to content

Instantly share code, notes, and snippets.

@angrycub
Created April 26, 2021 16:39
Show Gist options
  • Save angrycub/3a19f5fa18de8e0b4b40a24bbd462bd7 to your computer and use it in GitHub Desktop.
Save angrycub/3a19f5fa18de8e0b4b40a24bbd462bd7 to your computer and use it in GitHub Desktop.
Single Node Nomad Configuration

Single Node Nomad Configuration

This is a complex and unsupported configuration

Nomad is designed to be a highly-available, quorum-based cluster scheduler. Nomad deployments need to have three or more server nodes and one or more client nodes.

That said, Nomad does provide a dev mode which enables you to run Nomad with a single agent process. However, Nomad dev agents do behave slightly different than non-dev agents. To create a more authentic feeling experience, you might want to run Nomad with a configuration that runs both the client and server processes.

This configuration will enable you to run a single-node Nomad "cluster".

Running Nomad in this configuration comes with several tradeoffs and caveats, for example

  • The node has zero high availability -- the main drive for a scheduler.
  • If the state is damaged, you will have to wipe it and start over.
  • If your IP address changes, you have to perform peers.json recovery or wipe the state.
  • You can't run more than one instance of a workload that uses static ports.

If you are still determined to run Nomad this way, here are some annotated configuration files. Typically, I put them in /etc/nomad.d and start Nomad with nomad agent -config=/etc/nomad.d

@bobhenkel
Copy link

Do you have the annotated configuration files somewhere? Thanks

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment