I recently deployed a MAAS/Juju environment based on a six nodes cluster, using Ubuntu 12.04 LTS, to run a "proof of concept". I could appreciate how interesting the architecture is, and I'd like to understand if I can really base my future business on this technology. Therefore I'd like to understand more about any involved limitations, constraints and necessities.

In particular I'd like to understand how can MAAS be made High Availability, I.E. deployed over two or more physical nodes able to fail-over ?


HA is designed and planned for 14.04 in the case of both MAAS and Juju. You could manually construct it yourself now, but it would be rather difficult and then difficult to upgrade in future.

For now, I'd suggest using RAID mirroring and careful choice of nodes for reliability. If your MAAS or Juju go down you cannot scale up but you also don't lose the running services. When we deliver HA as part of the product it will be very easy to use and keep in sync with upstream, or Landscape will take care of it fully for you.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.