Hi all
I remember seeing a POC doc from someone on initial evaluation of Ceph. I would
like to see some examples and use cases I should go through. Is there such a
doc or blog post?
Thanks.___
ceph-users mailing list
ceph-users@lists.ceph.com
On 03/02/15 01:28, Loic Dachary wrote:
On 02/02/2015 13:27, Ritesh Raj Sarraf wrote:
By the way, I'm trying to build Ceph from master, on Ubuntu Trusty. I hope that
is supported ?
Yes, that's also what I have.
Same here - in the advent you need to rebuild the whole thing, using
This rbd_cache is only applicable to librbd not with the kernel rbd. Hope you
are trying with librbd based env.
If not, then the caching effect you are seeing is the filesystem cache.
Thanks Regards
Somnath
-Original Message-
From: ceph-users [mailto:ceph-users-boun...@lists.ceph.com]
I'm still missing something. I can check on the monitor to see that the running
config on the cluster has rbd cache = false
[root@essperf13 ceph]# ceph --admin-daemon
/var/run/ceph/ceph-mon.essperf13.asok config show | grep rbd
debug_rbd: 0\/5,
rbd_cache: false,
Since rbd caching is a
I've never seen this either. The build is slow and there is certainly
code reorg that could be done to speed it up but incremental builds
definitely work and are used extensively by all developers...
sage
On Mon, 2 Feb 2015, Loic Dachary wrote:
Hi,
I re-compile without cleaning and don't
On 02/02/2015 13:27, Ritesh Raj Sarraf wrote:
By the way, I'm trying to build Ceph from master, on Ubuntu Trusty. I hope
that is supported ?
Yes, that's also what I have.
On Mon, Feb 2, 2015 at 5:51 PM, Ritesh Raj Sarraf r...@researchut.com
mailto:r...@researchut.com wrote:
Thanks
By the way, I'm trying to build Ceph from master, on Ubuntu Trusty. I hope
that is supported ?
On Mon, Feb 2, 2015 at 5:51 PM, Ritesh Raj Sarraf r...@researchut.com
wrote:
Thanks Loic. I guess I need to look at the deb building script first then.
And now, looking at src/CMakeLists.txt, it is
While I don't know about this odd boost issue you see, personally I
find it useful to always be specific about built targets, as an
overall make will build e.g. unit tests -- quite slow. My favourite
is make ceph-mds ceph-osd ceph-mon ceph-fuse -- incremental builds
are pretty snappy that way.
Hi cephers,
I've got three questions:
1. Does anyone have an estimation on the release dates of the next stable ceph
branch?
2. Will the new stable release have improvements in the following areas: a)
working with ssd disks; b) cache tier
3. Will the new stable release introduce support
Thanks Loic. I guess I need to look at the deb building script first then.
And now, looking at src/CMakeLists.txt, it is clear that Ceph does make use
of reusing the build liraries.
On Mon, Feb 2, 2015 at 5:43 PM, Loic Dachary l...@dachary.org wrote:
Hi,
I re-compile without cleaning and
Hello Sudarshan,
Thanks, it should be useful when I want to appoint the specific OSD as
primary ;-)
On Mon, Feb 2, 2015 at 3:50 PM, Sudarshan Pathak sushan@gmail.com wrote:
Hello Dennis,
You can create CRUSH rule to select one of osd as primary as:
rule ssd-primary {
Hi,
We are currently working on adding changes to a sub-feature of Ceph.
My current challenge lies with the build environment of Ceph. Ceph is
huge and takes a lot of time to build. The build folder is close to 15 GiB.
I would like to re-use the compiled files when adding changes to the
Hi,
I re-compile without cleaning and don't have the same problem. It is supported
by Ceph, the problem is elsewhere.
My 2cts ;-)
On 02/02/2015 13:02, Ritesh Raj Sarraf wrote:
Hi,
We are currently working on adding changes to a sub-feature of Ceph.
My current challenge lies with the
On 30/01/15 13:39, Mark Kirkwood wrote:
On 30/01/15 12:34, Yehuda Sadeh wrote:
On Thu, Jan 29, 2015 at 3:27 PM, Mark Kirkwood
mark.kirkw...@catalyst.net.nz wrote:
On 30/01/15 11:08, Yehuda Sadeh wrote:
How does your regionmap look like? Is it updated correctly on all
zones?
Regionmap
It's not merely unstable, it's not actually complete. The XIOMessenger is
merged so that things don't get too far out of sync, but it should not be
used by anybody except developers who are working on it. :)
-Greg
On Mon, Feb 2, 2015 at 7:43 PM Nicheal zay11...@gmail.com wrote:
2015-02-03 0:48
Uh! It is strange. You said you have already cleared caches both on
client and osd node, the data must directly come from the disk. Wait
other's ideas
2015-02-03 11:44 GMT+08:00 Bruce McFarland bruce.mcfarl...@taec.toshiba.com:
Yes I'm using and the kernel rbd in Ubuntu 14.04 which makes calls
tried ceph on 3 kvm instances, each with a root 40G drive, and 6
virtio disks of 4G each. when i look at available space, instead of
some number less than 72G, i get 689G, and 154G used. the journal is
in a folder on the root drive. the images were made with virt-builder
using ubuntu-14.04 and
Hi,
Short of changing the kernel, we've not found away...
This is how we got it to work using the elrepo kernel-lt (3.10)
rpm -Uvh http://www.elrepo.org/elrepo-release-6-6.el6.elrepo.noarch.rpm
yum --enablerepo=elrepo-kernel install kernel-lt
sed -i 's/default=1/default=0/g' /etc/grub.conf
ceph 0.87
On Mon, Feb 2, 2015 at 7:53 PM, pixelfairy pixelfa...@gmail.com wrote:
tried ceph on 3 kvm instances, each with a root 40G drive, and 6
virtio disks of 4G each. when i look at available space, instead of
some number less than 72G, i get 689G, and 154G used. the journal is
in a
Hi Mika,
The below command will set ruleset to the pool:
ceph osd pool set poolname crush_ruleset 1
For more info : http://ceph.com/docs/master/rados/operations/crush-map/
Thanks
Sahana
From: ceph-users [mailto:ceph-users-boun...@lists.ceph.com] On Behalf Of Vickie
ch
Sent: Tuesday, February
On 02/03/2015 03:06 AM, Nick @ Deltaband wrote:
Hi Cephers,
If there is more than one public network, it possible to tell a client
(RBD) which public network to preference?
We have a working ceph cluster with a separate public (public1) and
cluster network. What i’d like to do is add a
On 3 February 2015 at 13:54, Wido den Hollander w...@42on.com wrote:
On 02/03/2015 03:06 AM, Nick @ Deltaband wrote:
Hi Cephers,
If there is more than one public network, it possible to tell a client
(RBD) which public network to preference?
We have a working ceph cluster with a separate
Hi ,
CRUSH map have two parameter are min_size and max_size.
Explanation about min_size is *If a pool makes fewer replicas than this
number, CRUSH will NOT select this rule*.
The max_size is *If a pool makes more replicas than this number, CRUSH
will NOT select this rule*
Default setting of
It seems you use the kernel rbd. So rbd_cache does not work, which is
just designed for librbd. Kernel rbd is directly using the system page
cache. You said that you have already run like echo 3
/proc/sys/vm/drop_cache to invalidate all pages cached in kernel. So
do you test the /dev/rbd1 based
Hi Christian,
On Fri, 30 Jan 2015 01:22:53 +0200 Georgios Dimitrakakis wrote:
Urged by a previous post by Mike Winfield where he suffered a
leveldb
loss
I would like to know which files are critical for CEPH operation
and
must
be backed-up regularly and how are you people doing it?
Hello!
We are planning to upgrade our Ubuntu 14.04.1 based cluster from Ceph
Firefly 0.80.7 to 0.80.8. We have 4 nodes, 12x4TB spinners each (plus
OS disks). Apart from the 12 OSDs per node, nodes 1-3 have MONs running.
The instructions on ceph.com say it is best to first restart the MONs,
There exists a more recently discuss in
PR(https://github.com/ceph/ceph/pull/1665).
On Mon, Feb 2, 2015 at 11:05 PM, J-P Methot jpmet...@gtcomm.net wrote:
Hi,
I've been looking into increasing the performance of my ceph cluster for
openstack that will be moved in production soon. It's a full
Hi all,
Does anyone using JCloud on Ceph?
Any feedback on the topic is welcome and will be very appreciated.
Regards
Alex
___
ceph-users mailing list
ceph-users@lists.ceph.com
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Hi,
I have a small, 3 node Firefly cluster. Each node hosting 6 OSDs, a 3 TB
spinner each. Each host has 2 SSDs used for the journals. Also each host
has 4 SSDs used as a 2 x RAID1 CacheCade array. The cluster is used to
host KVM based virutal machines, about 180 now. I'm thinking about
Hi,
I've been looking into increasing the performance of my ceph cluster for
openstack that will be moved in production soon. It's a full 1TB SSD
cluster with 16 OSD per node over 6 nodes.
As I searched for possible tweaks to implement, I stumbled upon
unitedstack's presentation at the
If you have K=2,M=1 you will survive one node failure. If your failure domain
is the host (i.e. there never is more than one chunk per node for any given
object), it will also survive two disks failures within a given node
because only one of them will have a chunk. It won't be able to resist
Hi Alexandre,
nice to meet you here ;-)
Hi Udo! (Udo from proxmox ? ;)
With 3 hosts only you can't survive an full node failure, because for
that you need
host = k + m.
And k:1 m:2 don't make any sense.
I start with 5 hosts and use k:3, m:2. In this case two hdds can fail or
one host
Are you actually using CMake? It's an alternative and incomplete build
system right now; the autotools build chain is the canonical one. (I don't
think it should be causing your problem, but...who knows?)
-Greg
On Mon, Feb 2, 2015 at 4:21 AM Ritesh Raj Sarraf r...@researchut.com wrote:
Thanks
Oh, yeah, that'll hurt on a small cluster more than a large one. I'm not
sure how much it matters, sorry.
On Mon, Feb 2, 2015 at 8:18 AM Daniel Schneller
daniel.schnel...@centerdevice.com wrote:
On 2015-02-02 16:09:27 +, Gregory Farnum said:
That said, for a point release it shouldn't
The packages might trigger restarts; the behavior has fluctuated a bit and
I don't know where it is right now.
That said, for a point release it shouldn't matter what order stuff gets
restarted in. I wouldn't worry about it. :)
-Greg
On Mon, Feb 2, 2015 at 6:47 AM Daniel Schneller
I mean it seemed fine for the current master branch ran under kernel
2.6.32 but I can't make sure that no other problem because no
production verify.
On Tue, Feb 3, 2015 at 12:21 AM, Haomai Wang haomaiw...@gmail.com wrote:
Hmm, I think there still some buggy exist in 2.6.32. I only try to
make
Hi Mad
3Gbps so you will have SSD Sata ?
I think you should take 6Gbps controllers to make sure so not have Sata
limitations
Thanks
Sent from my iPhone
On 2 févr. 2015, at 09:27, mad Engineer themadengin...@gmail.com wrote:
I am trying to create a 5 node cluster using 1 Tb SSD disks with 2
On Mon, Feb 2, 2015 at 5:27 AM, Andrei Mikhailovsky and...@arhont.com wrote:
Hi cephers,
I've got three questions:
1. Does anyone have an estimation on the release dates of the next stable
ceph branch?
We should be branching Hammer from master today, and it's
feature-frozen at this point. I
On 2015-02-02 16:09:27 +, Gregory Farnum said:
That said, for a point release it shouldn't matter what order stuff
gets restarted in. I wouldn't worry about it. :)
That is good to know. One follow-up then: If the packets trigger
restarts, they will most probably do so for *all* daemons
Thank you very much. Also thank you for the presentation you made in
Paris, it was very instructive.
So, from what I understand, the fiemap patch is proven to work on kernel
2.6.32 . The good news is that we use the same kernel in our setup. How
long have your production cluster been running
Hmm, I think there still some buggy exist in 2.6.32. I only try to
make write block size align(which already merged into master) but not
verify it in production. Our production cluster ran under customize
kernel version based on 3.12.
On Tue, Feb 3, 2015 at 12:18 AM, J-P Methot
On 02/02/2015 09:35 PM, Gregory Farnum wrote:
Are you actually using CMake? It's an alternative and incomplete build
system right now; the autotools build chain is the canonical one. (I
don't think it should be causing your problem, but...who knows?)
While CMake was installed, I doubt if it
Hi,
Writes will be distributed every 4MB (size of IMAGEV1 RBD object)
IMAGEV2 not fully supported on KRBD (but you can customize size of object and
striping)
You need to take :
- SSD SATA 6gbits
- or SSD SAS 12gbits (more expensive)
Florent Monthel
Le 2 févr. 2015 à 18:29, mad
Thanks Florent,
can ceph distribute write to multiple hosts?
On Mon, Feb 2, 2015 at 10:17 PM, Florent MONTHEL fmont...@flox-arts.net wrote:
Hi Mad
3Gbps so you will have SSD Sata ?
I think you should take 6Gbps controllers to make sure so not have Sata
limitations
Hi Cephers,
If there is more than one public network, it possible to tell a client
(RBD) which public network to preference?
We have a working ceph cluster with a separate public (public1) and
cluster network. What i’d like to do is add a new public (public2)
network for some newer clients. The
I'm using Ubuntu 14.04 and the kernel rbd which makes calls into libceph
root@essperf3:/etc/ceph# lsmod | grep rbd
rbd63707 1
libceph 225026 1 rbd
root@essperf3:/etc/ceph#
I'm doing raw device IO with either fio or vdbench (preferred tool) and there
is no
Yes I'm using and the kernel rbd in Ubuntu 14.04 which makes calls into libceph
root@essperf3:/etc/ceph# lsmod | grep rbd
rbd63707 1
libceph 225026 1 rbd
root@essperf3:/etc/ceph#
I'm doing raw device IO with either fio or vdbench (preferred tool) and there
2015-02-03 0:48 GMT+08:00 Gregory Farnum g...@gregs42.com:
On Mon, Feb 2, 2015 at 5:27 AM, Andrei Mikhailovsky and...@arhont.com wrote:
Hi cephers,
I've got three questions:
1. Does anyone have an estimation on the release dates of the next stable
ceph branch?
We should be branching
On Mon, Feb 2, 2015 at 11:28 AM, Andrei Mikhailovsky and...@arhont.com wrote:
I'm not sure what you mean about improvements for SSD disks, but the
OSD should be generally a bit faster. There are several cache tier
improvements included that should improve
I am trying to create a 5 node cluster using 1 Tb SSD disks with 2 OSD
on each server.Each server will have 10G NIC.
SSD disks are of good quality and as per label it can support ~300 MBps
What are the limiting factor that prevents from utilizing full speed
of SSD disks?
Disk controllers are 3
50 matches
Mail list logo