Having created a small single node Ceph cluster with following the 5 minute quickstart guide I was able to create a single node cluster with one OSD.
This probably wouldn’t be the first post that someone has written about this topic.
I’ve verified that it works in my test environment of Scientific Linux 6 by mounting the system with FUSE.
Here’s my fstab to describe my disk layout
# /etc/fstab # Created by anaconda on Fri Jul 6 14:27:56 2012 # # Accessible filesystems, by reference, are maintained under '/dev/disk' # See man pages fstab(5), findfs(8), mount(8) and/or blkid(8) for more info # /dev/mapper/vg_ceres-lv_root / ext4 defaults,user_xattr 1 1 UUID=4eb5efad-dbcd-4a9f-8187-d8ffa913e147 /boot ext4 defaults 1 2 /dev/mapper/vg_ceres-lv_home /data1 ext4 defaults,user_xattr 1 2 /dev/sdb /data ext4 defaults,user_xattr 1 2 /dev/mapper/vg_ceres-lv_swap swap swap defaults 0 0 tmpfs /dev/shm tmpfs defaults 0 0 devpts /dev/pts devpts gid=5,mode=620 0 0 sysfs /sys sysfs defaults 0 0 proc /proc proc defaults 0 0 Running a df -h gives this
After my last failed attempt at Installing Ceph on SL6 or rather my attempt at configuring Ceph for a test failed miserably.
It hasn’t deterred me to test more. As a result I setup a number of Vagrant Virtual machines and got together a few puppet scripts to provision machines.
Here’s a sample manifest for puppet to automate the deployment of a machine to build Ceph. It requires that you SL6 environment to have at least the epel repository enabled.
It’s somewhat interesting to listen to, we’ve been looking at ceph for a few things in work.
Just jotting this down so I remember to share this with the guys in work.
Over past few weeks I’ve been working on doing some integration and testing work to try and deliver a prototype system. I’ve taken the Vagrant tool and puppet to try and deliver systems for testing and development. Although the systems that I am currently working with aren’t fully automated, they are automated enough for them to be easily started up by a developer who reads the documentation (I hope).
I’m hoping that by providing these disposable systems that the various members in the team that I am working with will be more free to experiment and less fearful of breaking the system.
I’ve been a long time consumer of ruby applications but never quite got around to learning the language and the frameworks that are available to developers. Within the team that I have been working with, we’ve been evaluating hydra as a possible framework for our project.
I’ve been spending the last few days reading The RSpec Book, I’ve been able to pick up some of the basic syntax of the language from the BDD book.
Having discovered vagrant and veewee recently at OR2012 I’ve been building a few boxes for our site for testing purposes. I’ve had to relearn puppet to provision machines, but it’s paying off. I’ve been able to deploy 3-4 virtual machines with relative ease.
The only disappointment is the seemingly frequent and semi-random kernel panics and crashes that I get on OSX (which is probably caused by VirtualBox when I do lots of IO).
I only discovered Vagrant last week at OR2012 when one of the presentations had Vagrant as a part of the testing component in the project. I was pleasantly surprised that it was leaveraging lots of free and opensource technology to provide repeatable environments for development and testing. I was even more impressed by the Veewee project which provides some additional wrappers for building custom ‘boxes’. Veewee certainly saved me lots of time in setting up some updated Scientific Linux boxes (it’s my preferred distro for deployments).
I finally had gotten back from the conference late in the evening on Friday, airport security was not fun at all. It was the first Open Repositories conference that I had attended, the community seems to be pretty tightly knit. There were two or three software stacks which were presented that looked useful. Things such as Fedora-Commons, Hydra and a few other sub-components which were talked about are going to be likely building blocks that we will be using for our Digital Repository.
I only been here for a day and Edinburgh is as nice as I recall it was. This time my attendance to Open Repositories 2012 brought me and some co-workers of the DRI team to this city. It’s my first time at this event and there seems to be lots of people, so far it’s been pleasant enough apart from the weather.
For those interested in whats happening you should follow the hashtag #or2012 on twitter to see what people are saying.
As an exercise of a Friday afternoon of not starting something big before heading off to a conference. I’ve decided to spend an hour or two at seeing how ceph is installed and configured on an SL6 based machine (RHEL6 clone).
The install target is just an old desktop running a 64bit install of SL6x, so it’s nothing too fancy.
Following the instructions at http://ceph.com/wiki/Installing_on_RedHat_or_CentOS, I ended up doing this