pkalever / ceph-cluster

bringup a multi-node ceph cluster

ceph-cluster

Basic usage

---

To bring up a basic multi-node ceph cluster, just run:

$ vagrant up

The goal:

---

• Initiate a rbd block pool, create an image
  • In the summary, provide instructions to map image and mount the device
• Create ceph file system
  • In the summary, provide clear instructions to mount the CephFs volume

Dedicated ceph-admin node:

---

Make below changes to your Vagrantfile for a dedicated ceph admin node, else by default one of the ceph cluster nodes will be choosen to run admin tasks

Uncomment these lines
#config.vm.define "ceph-admin" do |node|
#  node.vm.hostname = "ceph-admin"
#  localoctet = 100+CEPH_NODES+1
#  node.vm.network :private_network, ip: "192.168.50.#{localoctet}"
#end

Also modify 's/ceph-node1/ceph-admin/' at below line
"ceph_admin" => ["ceph-node1"]

Get summary:

---

Here is an easy way to get the cluster details anytime

# ansible-playbook -i 192.168.50.101, ansible/summary.yml

Note: 192.168.50.101 is the static IP assigned to first ceph-node VM, no need to tweak it.