Container Image for ARM Architectures Available

Hello Celo community (and particularly Celo Node Operators)!

We are excited to announce the availability of a new container image for ARM architectures. This container image is now based on Debian, as opposed to our previous image, which was based on Alpine. You can use the same container tag for both linux/amd64 and linux/arm64 architectures, making it even more versatile for your deployment needs.

If you consider switching to an ARM 64 architecture, we believe it’ll be a good move for validators and node operators for the following reasons:

  • Increased Performance Efficency: Most ARM processors offer better speeds for the same energy consumption compared to AMD/Intel platforms.
  • Cost savings: Lower energy costs mean reduced operating costs, which results in more rewards being held as profit. Not to mention, lowering energy costs is in line with Celo’s mission.
  • Future proofing: Moving to an ARM architecture will ensure that you are compatible with emerging technologies, at least for a while.

Image Details:

  • Image Tag:
  • Source: Celo-Blockchain GitHub Repository (image automatically built with any commit to master brach)

At cLabs we have performed testing with arm64 container image, including full synchronization of the mainnet and testnets (alfajores and baklava). However, we strongly recommend that you test it in your specific environment before deploying it in a production setting. While we have taken every precaution to ensure its reliability, using it in your production environment is at your own risk.

The configuration and command flags are the same as for amd64 version. If someone using docker wants to run using a amd64, you can use docker run --platform linux/arm64 ....
Please let us know about your experiences with this image and any feedback you may have as it will help us continue to improve. We hope the image can help extending the different setups and infrastructure offering where the community can run celo nodes.

If you do decide to make the switch to an ARM architecture, please share your experiences here so others can benefit as well.

Happy building and testing!

Best regards,


Running a full node on a raspberry pi 4 using the new image.

Works as expected! Even a little faster now that I don’t have to use an emulator to run the node on the RPi