Ceph Installation :: Part-1


Ceph Installation Step by Step

This quick start setup helps to deploy ceph with 3 Monitors and 2 OSD nodes with 4 OSD each node. In this we are using commodity hardware running CentOS 6.4

Ceph-mon1 : First Monitor + Ceph-deploy machine (will be used to deploy ceph to other nodes )

Ceph-mon2 : Second Monitor ( for monitor quorum )

Ceph-mon3 : Third Monitor ( for monitor quorum )

Ceph-node1 : OSD node 1 with 10G X 1 for OS , 440G X 4 for 4 OSD

Ceph-node2 : OSD node 2 with 10G X 1 for OS , 440G X 4 for 4 OSD

Ceph-Deploy Version is 1.3.2 , Ceph Version 0.67.4 ( Dumpling )

Preflight Checklist 

All the Ceph Nodes may require some basic configuration work prior to deploying a Ceph Storage Cluster.

CEPH node setup

  • Create a user on each Ceph Node.

sudo useradd -d /home/ceph -m ceph
sudo passwd ceph

  • Add root privileges for the user on each Ceph Node.

echo "ceph ALL = (root) NOPASSWD:ALL" | sudo tee /etc/sudoers
sudo chmod 0440 /etc/sudoers

  • Configure your ceph-deploy node ( ceph-mon1) with password-less SSH access to each Ceph Node. Leave the passphrase empty , repeat this step for CEPH and ROOT users.

ceph@ceph-admin:~ [ceph@ceph-admin ~]$ ssh-keygen
Generating public/private rsa key pair.
Enter file in which to save the key (/home/ceph/.ssh/id_rsa): yes
Created directory '/home/ceph/.ssh'.
Enter passphrase (empty for no passphrase):
Enter same passphrase again:
Your identification has been saved in /home/ceph/.ssh/id_rsa.
Your public key has been saved in /home/ceph/.ssh/id_rsa.pub.
The key fingerprint is:
48:86:ff:4e:ab:c3:f6:cb:7f:ba:46:33:10:e6:22:52 ceph@ceph-admin.csc.fi
The key's randomart image is:
+--[ RSA 2048]----+
| |
| E. o |
| .. oo . |
| . .+..o |
| . .o.S. |
| . + |
| . o. o |
| ++ .. . |
| ..+*+++ |

  • Copy the key to each Ceph Node. ( Repeat this step for ceph and root users )

[ceph@ceph-mon1 ~]$ ssh-copy-id ceph@ceph-node2
The authenticity of host 'ceph-node2 (' can't be established.
RSA key fingerprint is ac:31:6f:e7:bb:ed:f1:18:9e:6e:42:cc:48:74:8e:7b.
Are you sure you want to continue connecting (yes/no)? y
Please type 'yes' or 'no': yes
Warning: Permanently added 'ceph-node2,' (RSA) to the list of known hosts.
ceph@ceph-node2's password:
Now try logging into the machine, with "ssh 'ceph@ceph-node2'", and check in: .ssh/authorized_keys
to make sure we haven't added extra keys that you weren't expecting.
[ceph@ceph-mon1 ~]$

  • Ensure connectivity using ping with hostnames , for convenience we have used local host file , update host file of every node with details of other nodes. PS : Use of DNS is recommended
  • Packages are cryptographically signed with the release.asc key. Add release key to your system’s list of trusted keys to avoid a security warning:

sudo rpm --import 'https://ceph.com/git/?p=ceph.git;a=blob_plain;f=keys/release.asc'

  • Ceph may require additional additional third party libraries. To add the EPEL repository, execute the following:

su -c 'rpm -Uvh http://dl.fedoraproject.org/pub/epel/6/x86_64/epel-release-6-8.noarch.rpm'
sudo yum install snappy leveldb gdisk python-argparse gperftools-libs

  • Installing Release packages , Dumpling is the most recent stable release of Ceph. ( by the time i am creating this wiki )

su -c 'rpm -Uvh http://ceph.com/rpm-dumpling/el6/noarch/ceph-release-1-0.el6.noarch.rpm'

  • Adding ceph to YUM , create repository file for ceph /etc/yum.repos.d/ceph.repo

name=Ceph packages for $basearch

name=Ceph noarch packages

name=Ceph source packages

  • For best results, create directories on your nodes for maintaining the configuration generated by ceph . These should get auto created by ceph however in may case it gave me problems. So creating manually.

mkdir -p /etc/ceph /var/lib/ceph/{tmp,mon,mds,bootstrap-osd} /var/log/ceph

  • By default, daemons bind to ports within the 6800:7100 range. You may configure this range at your discretion. Before configuring your IP tables, check the default iptables configuration. ::ports within the 6800:7100 range. You may configure this range at your discretion. Since we are performing test deployment we can disable iptables on ceph nodes . For moving to production this need to be attended.

Please Follow Ceph Installation :: Part-2 for next step in installation