I did it mostly to learn about docker (and because i wanted a 'cluster
in the office' like Loic had :-)). Not incredibly pretty though.
Usually i go with vagrant to deploy ceph in a box (1 mon, 2 osds) in
an lxc container, with file loopback mounts and using the puppet
module to deploy/config (whi
The packer solution sounds sensible to me as it also would allow the
creation of Docker images.
A basic Docker image has already been started
(https://index.docker.io/u/rochaporto/cephbox/) but it would be quite
useful to have a version that comes from an official maintained
source.
On Tue, Feb 2
> so that’s great. It’s already almost done. Let us know when you have managed
> updating it.
>
> I suppose it would be best if we could offer one image each of the last 3
> Ceph versions (dumpling, emperor and firefly) so people can then play with
> the expected level they are looking for in ei
Hi Sebastien,
so that’s great. It’s already almost done. Let us know when you have managed
updating it.
I suppose it would be best if we could offer one image each of the last 3 Ceph
versions (dumpling, emperor and firefly) so people can then play with the
expected level they are looking for
Hi Patrick,
This is great news. Not only for the students who will get a chance to discover
Ceph in the process. But also because it means you managed to demonstrate that
Ceph is a project supported by a community.
Congrats :-)
On 24/02/2014 21:42, Patrick McGarry wrote:
> Hey Cephers,
>
> T
FYI I already have a vagrant box ready.
http://www.sebastien-han.fr/blog/2013/04/22/play-with-ceph-vagrant-box/
The box just needs to be updated with the last version.
Cheers.
Sébastien Han
Cloud Engineer
"Always give 100%. Unless you're giving blood.”
Phone: +33 (0)1 49 70 99 72
Mai
On 02/24/2014 04:33 PM, Jean-Charles LOPEZ wrote:
Hi Joe,
I already have a VM that is very close to what Sage is talking about.
Just need to get the MDS configured on it and it should be ready.
Will keep you in the loop when I need scripting it.
Ok. Thanks.
--
Joseph Landman, Ph.D
Founder
On 02/24/2014 04:26 PM, Loic Dachary wrote:
On 24/02/2014 21:59, Jean-Charles LOPEZ wrote:
Hi Sage,
I think this is a great idea.
However, I would recommend that two options to be available: 1)
First options is a standalone package as you describe it 2) Second
option is a prepackaged VM in O
On 24/02/2014 21:59, Jean-Charles LOPEZ wrote:
> Hi Sage,
>
> I think this is a great idea.
>
> However, I would recommend that two options to be available:
> 1) First options is a standalone package as you describe it
> 2) Second option is a prepackaged VM in OVA format people can just downloa
Hi Sage,
I think this is a great idea.
However, I would recommend that two options to be available:
1) First options is a standalone package as you describe it
2) Second option is a prepackaged VM in OVA format people can just download and
import using VirtualBox, Fusion or VMWarePlayer.
I can
Hey Cephers,
This week we found out that the Ceph project has been accepted as a
mentoring organization [0] for the Google Summer of Code [1]. This
means that Google will foot the bill for some number of student
developers this summer ($6000, of which $5500 goes to the student and
$500 goes to Ce
-- All Branches --
Alfredo Deza
2013-09-27 10:33:52 -0400 wip-5900
2014-01-21 16:29:22 -0500 wip-6465
Dan Mick
2013-07-16 23:00:06 -0700 wip-5634
2014-02-12 23:29:28 -0800 wip-no-version
2014-02-18 13:07:32 -0800 wip-7463
Danny Al-Gaaf
On 02/21/2014 12:55 PM, Ilya Dryomov wrote:
> In preparation for prefixing rbd writes with an allocation hint
> introduce a num_ops parameter for rbd_osd_req_create(). The rationale
> is that not every write request is a write op that needs to be prefixed
> (e.g. watch op), so the num_ops logic ne
On 02/21/2014 12:55 PM, Ilya Dryomov wrote:
> CEPH_OSD_MAX_OP value in rbd_osd_req_callback() is hard-coded to 2.
> Fix it.
Please squash this in with the previous patch (at least).
Change the BUG_ON() to rbd_assert() while you're at it
(and invert the logic).
Reviewed-by: Alex Elder
> Signed-o
On 02/21/2014 12:55 PM, Ilya Dryomov wrote:
> In an effort to reduce fragmentation, prefix every rbd write with
> a CEPH_OSD_OP_SETALLOCHINT osd op with an expected_write_size value set
> to the object size (1 << order). Backwards compatibility is taken care
> of on the libceph/osd side.
If *ever
On 02/21/2014 12:55 PM, Ilya Dryomov wrote:
> Our longest osd request now contains 3 ops: copyup+hint+write.
This is pretty trivial and could be squashed into the
patch that actually starts using more than two ops.
It could certainly be squashed together with the next
patch (that fixes the asserti
On 02/21/2014 12:55 PM, Ilya Dryomov wrote:
> This is primarily for rbd's benefit and is supposed to combat
> fragmentation:
>
> "... knowing that rbd images have a 4m size, librbd can pass a hint
> that will let the osd do the xfs allocation size ioctl on new files so
> that they are allocated in
On 02/21/2014 12:55 PM, Ilya Dryomov wrote:
> Hello,
>
> This series adds support for CEPH_OSD_OP_SETALLOCHINT osd op to libceph
> along with adjusting rbd to make use of it. The rationale and the
> basic desing was outlined in the "rados io hints" thread on ceph-devel
> about a month ago.
It lo
On 02/21/2014 12:55 PM, Ilya Dryomov wrote:
> Encode ceph_osd_op::flags field so that it gets sent over the wire.
This looks fine. The flags field was never being encoded
before, so there's no change in the protocol. And you're
just adding a flags field to the local structure.
Reviewed-by: Alex
> and have a usable mini-cluster. That makes it really easy for users to play
> with the basics, like:
[Martin Maurer]
Hello Sage,
Not the same, but another fast way to get a full OS and Ceph installed for
testing (and more): Proxmox VE Ceph Server
For playing with Ceph, you can install Pro
I think it would be interesting.
I am actually using a shell script written by Loic Dachary from his blog
(http://dachary.org/?p=2374).
The ceph-standalone package would be very useful for fast appropriation and
understanding of Ceph.
This kind of installation can facilitate client functional
Hi Eric,
Great !! The cluster is totally up now
I got a wrong key in ceph.bootstrap-osd.keyring file in /etc/ceph of the
ceph-deploy server.
De : eric mourgaya [mailto:eric.mourg...@gmail.com]
Envoyé : vendredi 21 février 2014 15:39
À : CHEVALIER Ghislain IMT/OLPS
Cc : ceph-devel@vger.kernel
22 matches
Mail list logo