Skip to content
This repository

HTTPS clone URL

Subversion checkout URL

You can clone with HTTPS or Subversion.

Download ZIP

Ceph cookbook

branch: master

Fetching latest commit…

Octocat-spinner-32-eaf2f5

Cannot retrieve the latest commit at this time

Octocat-spinner-32 attributes
Octocat-spinner-32 libraries
Octocat-spinner-32 providers
Octocat-spinner-32 recipes
Octocat-spinner-32 resources
Octocat-spinner-32 templates
Octocat-spinner-32 README.rdoc
Octocat-spinner-32 metadata.rb
README.rdoc

DESCRIPTION

This cookbook will configure a set of machines to act as a Ceph cluster. It is intended to be as easy as possible to configure, and will take a lot of sane defaults to start with.

Currently, there are some limitations:

  • It is only tested on Debian Squeeze 64 bit - but might work on other distro's

  • It depends on the apt-cookbook from Opscode

  • It only configures mons and osds

CONTENTS

There are LWRPs for MON, OSD, KEYRING and CONFIG (the MDS is present but non-functional).

USAGE

General

  1. Add this cookbook to your tree, and name it “ceph”. We do this as a submodule:

    git submodule add user@host:/where/you/pull/from cookbooks/ceph
  2. Upload the cookbook to your chef server.

  3. Create a role (“cephcluster”), and set at least this attribute:

    [:ceph][:clustername] - set it to a global identifier for this cluster
    [:ceph][:mons] - set it to an array of fqdn names for the mons

    Apply this role to all your clusternodes so the attributes are know to each node.

Monitoring cluster

  1. Apply the “ceph::mon” recipe to a monitor node, run chef-client there, it should become the first monitor, and also the “master” which holds the client.admin secret.

  2. Run ceph::mon recipe on all other nodes which are mons

  3. Verify by running “ceph -w” on any mon-node

Adding OSDs

OSDs are identified by the LABEL that sits on certain partition or drive. Each device that might be used as a OSD should have this label. Format the block devices (either manual or by creating your own chef recipe) and add a label named “{clustername}.ceph”.

mkfs.btrfs -L ceph.ceph /dev/sdb
mkfs.btrfs -L ceph.ceph /dev/sdc
mkfs.btrfs -L ceph.ceph /dev/sdd

Once done, add the OSD recipe to the node, and run chef-client. The labels will be detected, and the devices will be initialized and started. If the device holds an already initialized osd-structure, it will not re-initialize.

The OSD daemons are normally started.

IMPROVEMENTS:

This is a work in progress, and anyone is invited to contribute their changes/additions.

If you want to work on this, and don't know where to start:

  • mds (LWRP)

  • radosgw setup (LWRP)

  • rbd management through a LWRP

Get in touch on IRC wonko (kreynet) or wonko_be (oftc).

LICENSE and AUTHOR:

Author: Bernard Grymonpon (bernard@openminds.be) Author: Openminds BVBA devops team (info@openminds.be)

Copyright 2009-2012 Openminds BVBA.

Licensed under the Apache License, Version 2.0 (the “License”); you may not use this file except in compliance with the License. You may obtain a copy of the License at

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an “AS IS” BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.

Something went wrong with that request. Please try again.