Hi! Recently I took administration of an existing 3-node PVE/Ceph Cluster. First task/project is to plan a relocation of a small office branch datacentre. There is some network equipment, few physical servers and the PVE/Ceph cluster. Most of this is easy stuff, Ceph is quite new. Yeah!
As I am currently preparing a process for this I wondered how to deal with the PVE especially the Ceph-part of the cluster. It is a three node cluster, 12 OSD each, with a dedicated 50GbE-link. The network and IP configuration at the other datacentre keeps the same. So this should be fine.
My plan so far for a weekend maintenance trip (updating the plan here continously with information gathered):
I have dealt with proxmox pve quite some time but I am relatively new to this Ceph stuff. Is there anything I am obviously missing, mostly regarding Ceph?
Thanks for any help with this
References:
As I am currently preparing a process for this I wondered how to deal with the PVE especially the Ceph-part of the cluster. It is a three node cluster, 12 OSD each, with a dedicated 50GbE-link. The network and IP configuration at the other datacentre keeps the same. So this should be fine.
My plan so far for a weekend maintenance trip (updating the plan here continously with information gathered):
- Shutdown all VM/CT on all nodes. Check they are/stay down.
- Evaluate requiered / optional CEPH flags like "noout" - which are necessary, which are optional...
- Shutdown all PVE cluster nodes.
- Relocate all equipment (by car - drink some tea and listen to some good music on the road)
- At the target site: Setup boxes and cabling
- Start all cluster nodes.
- Hopefully seeing all cluster nodes and the CEPH status going all green
- Starting the VM/CTs
- Check everything.
- Re-check everything.
- Drive back home. (& later enjoy a beer at home)
I have dealt with proxmox pve quite some time but I am relatively new to this Ceph stuff. Is there anything I am obviously missing, mostly regarding Ceph?
Thanks for any help with this
References:
- [1] https://forum.proxmox.com/threads/ceph-wont-properly-start-after-maintenance.100533/
- [2] https://forum.proxmox.com/threads/w...and-upgrade-to-pve-and-ceph.56082/post-260144
- [3] https://forum.proxmox.com/threads/w...and-upgrade-to-pve-and-ceph.56082/post-260165
- [4] https://pve.proxmox.com/wiki/Ceph_Octopus_to_Pacific#Upgrade_on_each_Ceph_cluster_node
- [5] https://forum.proxmox.com/threads/shutdown-of-the-hyper-converged-cluster-ceph.68085/
- [6] https://forum.proxmox.com/threads/shutdown-of-the-hyper-converged-cluster-ceph.68085/post-395062
- [7] https://old.ceph.com/planet/how-to-do-a-ceph-cluster-maintenance-shutdown/
Last edited: