Hello dear family, happy year 2021, success in your goals this year. I would like to know, what would be the correct order of activities to carry out in case the frontend server, the nodes and VMs that it has running were to shut down in case of a power failure. That is, how to restore the services in order and correctly, I know that Opennebula works with demons and sometimes there are processes that do not end well and infer at the time of restart. I would appreciate the prompt response. Cheers
I would also welcome a guide from the OpenNebula team for the best way to start/stop the entire stack.
We are running in 5.6 and 5.12 versions. Below are the sequence that we are following to bring down and bring up the stack during maintenance:
Bring down sequence:
- Gracefully power off all the VMs running in sunstone
- Disable the nodes in sunstone
- Gracefully power off the nodes
- Gracefully power off the frontend server (It is recommended to take the DB backup before powering off the frontend server)
- Gracefully power off the shared storage
Bring up sequence:
- Turn on the shared storage
- Turn on the nodes (Ensure that the libvirt services are running)
- Turn on the frontend server (Ensure that the OpenNebula and OpenNebula sunstone services are running)
- Check if the storage is mounted in all the nodes and frontend server, if you are using any shared storages.
- Enable the nodes in sunstone.
- Power on the VMs one by one in sunstone
Hopefully this will help you. Thank you.