Project

General

Profile

Ceph Vagrant Setup

[Reposted from: Sebastien Han]

Materials to start playing with Ceph. This Vagrant box contains a all-in-one Ceph installation.

I. SETUP

First Download and Install Vagrant.
Download the Ceph box: here. This box contains one virtual machine:
  • Ceph VM contains 2 OSDs (1 disk each), 1 MDS, 1 MON, 1 RGW. A modified CRUSH Map, it simply represents a full datacenter and applies a replica per OSD
  • VagrantFile for both VM client and ceph
  • Other include files
Download an extra VM for the client here, note that Debian and Red Hat based system work perfectly, thus it’s up to you:
  • Client: just an Ubuntu installation

Initialize the Ubuntu box:

$ wget https://www.dropbox.com/s/hn28qgjn59nud6h/ceph-all-in-one.box
...
...
$ vagrant box add big-ceph ceph-all-in-one.box
[vagrant] Downloading with Vagrant::Downloaders::File...
[vagrant] Copying box to temporary location...
[vagrant] Extracting box...
[vagrant] Verifying box...
[vagrant] Cleaning up downloaded box...

Initialize the Client box:

$ wget http://dl.dropbox.com/u/1537815/precise64.box
$ vagrant box add ubuntu-12.04.1 precise64.box
[vagrant] Downloading with Vagrant::Downloaders::File...
[vagrant] Copying box to temporary location...
[vagrant] Extracting box...
[vagrant] Verifying box...
[vagrant] Cleaning up downloaded box...

Check your boxes:

$ vagrant box list
ceph-all-in-one
ubuntu-12.04.1

Import all the files from the box:

 
$ mkdir setup
$ cp /Users/leseb/.vagrant.d/boxes/ceph-all-in-one/include/* setup/
$ mv setup/_Vagrantfile Vagrantfile

In order to make the setup easy, I assume that your working directory is $HOME/ceph. At the end, your tree directory looks like this:

.
??? Vagrantfile
??? ceph-all-in-one.box
??? precise64.box
??? setup
    ??? ceph.conf
    ??? ceph.sh
    ??? keyring

II. START IT!

Check the state of your virtual machines:

$ vagrant status
Current VM states:
client                   poweroff
ceph                     poweroff
This environment represents multiple VMs. The VMs are all listed
above with their current state. For more information about a specific
VM, run `vagrant status NAME`.

Eventually run them:

$ vagrant up ceph && vagrant up client
...
...

The next time, you’ll run the client, run it this way to don’t re-provision the machine:

$ vagrant up --no-provision client

Eventually SSH on your client:

$ vagrant ssh client
...
vagrant@ceph:~$ sudo ceph -s   
   health HEALTH_OK
   monmap e3: 1 mons at {1=192.168.251.100:6790/0}, election epoch 1, quorum 0 1
   osdmap e179: 2 osds: 2 up, 2 in
   pgmap v724: 96 pgs: 96 active+clean; 9199 bytes data, 2071 MB used, 17906 MB / 19978 MB avail; 232B/s wr, 0op/s
   mdsmap e54: 1/1/1 up {0=0=up:active}
vagrant@ceph:~$ sudo ceph osd tree
# id  weight  type name up/down reweight
-1  2 root default
-4  2   datacenter dc
-5  2     room laroom
-6  2       row larow
-3  2         rack lerack
-2  2           host ceph
0 1             osd.0 up  1
1 1             osd.1 up  1

III. BONUS UPGRADE TO CUTTLEFISH

It’s fairly easy to upgrade the box to last stable version Cuttlefish. For this simply edit /etc/apt/sources.list.d/ceph.list/ceph.list with the following:

deb http://ceph.com/debian-cuttlefish/ precise main

Then run:

$ sudo apt-get update && apt-get install ceph
$ sudo service ceph restart
$ sudo ceph -v
ceph version 0.61 (237f3f1e8d8c3b85666529860285dcdffdeda4c5)

Note: if for some reasons you get a status were only 1/2 OSDs are up, just restart the mon. This should do the trick :-).