Hi Yan-
Thanks! Tracked this down to teh Pipe AuthSessionHandler. Fixed in
bcefc0e80ac2bda6d3b4550a6d849302bcc201d6.
sage
On Sat, 3 Nov 2012, Yan, Zheng wrote:
Hi Everybody,
I found there is a memory leak bug in the latest master (c51e1f9b64).
The bug affects
both ceph-osd and
Hello,
As far I'm concerned I think that 12 disks per servers is way too much.
From your experience, what is the correct number of OSD per server ?
Denis
--
To unsubscribe from this list: send the line unsubscribe ceph-devel in
the body of a message to majord...@vger.kernel.org
More
On Sun, Nov 4, 2012 at 10:58 AM, Aleksey Samarin nrg3...@gmail.com wrote:
Hi all
Im planning use ceph for cloud storage.
My test setup is 2 servers connected via infiniband 40Gb, 6x2Tb disks per
node.
Centos 6.2
Ceph 0.52 from http://ceph.com/rpms/el6/x86_64
This is my config
Hi Yan,
On Sat, 3 Nov 2012, Yan, Zheng wrote:
From: Yan, Zheng zheng.z@intel.com
ceph_aio_write() has an optimization that marks cap EPH_CAP_FILE_WR
dirty before data is copied to page cache and inode size is updated.
If sceph_check_caps() flushes the dirty cap before the inode size is
The wip-explicit-pgnum changes the 'ceph osd pool create name pgnum'
command to require the pg_num value instead of defaulting to 8. This
would make it harder for users to get this wrong.
On the other hand, it probably also breaks some scripts for deploying
OpenStack that create volume and
On Sun, Nov 4, 2012 at 1:04 PM, Aleksey Samarin nrg3...@gmail.com wrote:
Hi!
This command? ceph tell osd \* bench
Output: tell target 'osd' not a valid entity name
Well, i did pool by command ceph osd pool create bench2 120
This output of rados -p bench2 bench 30 write --no-cleanup
rados
On Sat, 3 Nov 2012, Noah Watkins wrote:
The libcephfs interface asserts valid file descriptors, so clients are
forced to manage this or risk crashes. Pushed to wip-bad-fd is changes
that handle this case by returning -EBADF. Review welcome :)
This looks good to me. The one thing I'd change
[Sorry for the blank email; I missed!]
On Sun, Nov 4, 2012 at 1:04 PM, Aleksey Samarin nrg3...@gmail.com wrote:
Hi!
This command? ceph tell osd \* bench
Output: tell target 'osd' not a valid entity name
I guess it's ceph osd tell \* bench. Try that one. :)
Well, i did pool by command ceph
It`s ok!
Output:
2012-11-04 16:19:23.195891 osd.0 [INF] bench: wrote 1024 MB in blocks
of 4096 KB in 11.441035 sec at 91650 KB/sec
2012-11-04 16:19:24.981631 osd.1 [INF] bench: wrote 1024 MB in blocks
of 4096 KB in 13.225048 sec at 79287 KB/sec
2012-11-04 16:19:25.672896 osd.2 [INF] bench: wrote
On 11/04/2012 03:58 AM, Aleksey Samarin wrote:
Hi all
Im planning use ceph for cloud storage.
My test setup is 2 servers connected via infiniband 40Gb, 6x2Tb disks per node.
Centos 6.2
Ceph 0.52 from http://ceph.com/rpms/el6/x86_64
This is my config http://pastebin.com/Pzxafnsm
One thing that
That's only nine — where are the other three? If you have three slow
disks that could definitely cause the troubles you're seeing.
Also, what Mark said about sync versus syncfs.
On Sun, Nov 4, 2012 at 1:26 PM, Aleksey Samarin nrg3...@gmail.com wrote:
It`s ok!
Output:
2012-11-04
Ok!
Well, I'll take these tests and write about the results.
btw,
disks are the same, as some may be faster than others?
2012/11/4 Gregory Farnum g...@inktank.com:
That's only nine — where are the other three? If you have three slow
disks that could definitely cause the troubles you're seeing.
Well, i create ceph cluster with 2 osd ( 1 osd per node), 2 mon, 2 mds.
here is what I did:
ceph osd pool create bench
ceph osd tell \* bench
rados -p bench bench 30 write --no-cleanup
output:
Maintaining 16 concurrent writes of 4194304 bytes for at least 30 seconds.
Object prefix:
Hi,
Does it mean that we should destroy those pool and recreate them with higher
pgnum ?
--
Yann
-Message d'origine-
De : ceph-devel-ow...@vger.kernel.org [mailto:ceph-devel-ow...@vger.kernel.org]
De la part de Yehuda Sadeh
Envoyé : vendredi 2 novembre 2012 20:20
À : Sylvain Munaut
On Sun, Nov 4, 2012 at 7:29 PM, Sage Weil s...@inktank.com wrote:
Hi Yan,
On Sat, 3 Nov 2012, Yan, Zheng wrote:
From: Yan, Zheng zheng.z@intel.com
ceph_aio_write() has an optimization that marks cap EPH_CAP_FILE_WR
dirty before data is copied to page cache and inode size is updated.
On Fri, 2 Nov 2012, Nick Bartos wrote:
Sage,
A while back you gave us a small kernel hack which allowed us to mount
the underlying OSD xfs filesystems in a way that they would ignore
system wide syncs (kernel hack + mounting with the reused mand
option), to workaround a deadlock problem
On 11/04/2012 07:18 AM, Aleksey Samarin wrote:
Well, i create ceph cluster with 2 osd ( 1 osd per node), 2 mon, 2 mds.
here is what I did:
ceph osd pool create bench
ceph osd tell \* bench
rados -p bench bench 30 write --no-cleanup
output:
Maintaining 16 concurrent writes of 4194304
On Sun, 4 Nov 2012, Yan, Zheng wrote:
On Sun, Nov 4, 2012 at 7:29 PM, Sage Weil s...@inktank.com wrote:
Hi Yan,
On Sat, 3 Nov 2012, Yan, Zheng wrote:
From: Yan, Zheng zheng.z@intel.com
ceph_aio_write() has an optimization that marks cap EPH_CAP_FILE_WR
dirty before data is
On Sun, 4 Nov 2012, Stefan Priebe wrote:
Hello,
while benchmarking the new v2 rbd format i saw several crashes of KVM. First i
thought this was kernel related but i was just happening faster with one
kernel than the other.
Can you try the wip-rbd-read branch? It has a bunch of fixes that
Can i merge wip-rbd-read into master?
Stefan
Am 04.11.2012 15:06, schrieb Sage Weil:
On Sun, 4 Nov 2012, Stefan Priebe wrote:
Hello,
while benchmarking the new v2 rbd format i saw several crashes of KVM. First i
thought this was kernel related but i was just happening faster with one
kernel
On Sun, 4 Nov 2012, Stefan Priebe wrote:
Can i merge wip-rbd-read into master?
Yeah. I'm going to do a bit more testing first before I do it, but it
should apply cleanly. Hopefully later today.
sage
Stefan
Am 04.11.2012 15:06, schrieb Sage Weil:
On Sun, 4 Nov 2012, Stefan Priebe
On Sun, Nov 4, 2012 at 10:01 PM, Sage Weil s...@inktank.com wrote:
On Sun, 4 Nov 2012, Yan, Zheng wrote:
On Sun, Nov 4, 2012 at 7:29 PM, Sage Weil s...@inktank.com wrote:
Hi Yan,
On Sat, 3 Nov 2012, Yan, Zheng wrote:
From: Yan, Zheng zheng.z@intel.com
ceph_aio_write() has an
Did your RAID setup improve anything?
I have tried to launch 2 fio test in parallel, on 2 disks in the same guest
vm, I get 2500iops for each test
Running 2 fio tests, on 2 differents guests, give me 5000iops for each test.
I really don't understand...Maybe something don't use parallelim
On Sun, Nov 4, 2012 at 4:17 AM, Sage Weil s...@inktank.com wrote:
but this is actually doing the lookup twice. A better pattern:
mapfoo, bar::iterator p = container.find(key);
if (p == container.end())
return NULL;
return p-second;
Cool, thanks for
What may be possible solutions?
Update centos to 6.3?
About issue with writes to lots of disk, i think parallel dd command
will be good as test! :)
2012/11/4 Mark Nelson mark.nel...@inktank.com:
On 11/04/2012 07:18 AM, Aleksey Samarin wrote:
Well, i create ceph cluster with 2 osd ( 1 osd per
Well, there is not really an ideal number. For instance, at
Dreamhost they put 12 disks per machines, but they have really big and
powerful server. The problem is: as many disks per machine you have
as data you have to migrate in case of failure. At the end, it's up to
you... It also highly
On Sun, 4 Nov 2012, Yan, Zheng wrote:
Short write happens when we fail to get 'Fb' cap for all pages. Why
shouldn't
we fall back to sync write, I think some user programs assume short write
never happen unless ENOSPC. If generic_file_aio_write return and its return
value shows there was
Hello,
does the mon service also suffer from btrfs as fs? Right now i've only
the osd's running btrfs.
Greets,
Stefan
--
To unsubscribe from this list: send the line unsubscribe ceph-devel in
the body of a message to majord...@vger.kernel.org
More majordomo info at
On Sun, 4 Nov 2012, Stefan Priebe wrote:
Am 04.11.2012 15:12, schrieb Sage Weil:
On Sun, 4 Nov 2012, Stefan Priebe wrote:
Can i merge wip-rbd-read into master?
Yeah. I'm going to do a bit more testing first before I do it, but it
should apply cleanly. Hopefully later today.
Any
Am 04.11.2012 20:19, schrieb Sage Weil:
On Sun, 4 Nov 2012, Stefan Priebe wrote:
Am 04.11.2012 15:12, schrieb Sage Weil:
On Sun, 4 Nov 2012, Stefan Priebe wrote:
Can i merge wip-rbd-read into master?
Yeah. I'm going to do a bit more testing first before I do it, but it
should apply
On Sun, 4 Nov 2012, Stefan Priebe wrote:
Am 04.11.2012 20:19, schrieb Sage Weil:
On Sun, 4 Nov 2012, Stefan Priebe wrote:
Am 04.11.2012 15:12, schrieb Sage Weil:
On Sun, 4 Nov 2012, Stefan Priebe wrote:
Can i merge wip-rbd-read into master?
Yeah. I'm going to do a bit more
Awesome, thanks! I'll let you know how it goes.
On Sun, Nov 4, 2012 at 5:50 AM, Sage Weil s...@inktank.com wrote:
On Fri, 2 Nov 2012, Nick Bartos wrote:
Sage,
A while back you gave us a small kernel hack which allowed us to mount
the underlying OSD xfs filesystems in a way that they would
On 16-10-12 19:14, James Page wrote:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA256
Hi All
I've started to put together a Launchpad Blueprint to act as a
placeholder for discussion at the Ubuntu Developer Summit in a couple
of weeks time.
I'd be interested in feedback on what folks would
On Mon, Nov 5, 2012 at 12:45 AM, Sage Weil s...@inktank.com wrote:
On Sun, 4 Nov 2012, Yan, Zheng wrote:
Short write happens when we fail to get 'Fb' cap for all pages. Why
shouldn't
we fall back to sync write, I think some user programs assume short write
never happen unless ENOSPC. If
Unfortunately I'm still seeing deadlocks. The trace was taken after a
'sync' from the command line was hung for a couple minutes.
There was only one debug message (one fs on the system was mounted with 'mand'):
kernel: [11441.168954] [8113538a] ? sync_fs_one_sb+0x4d/0x4d
Here's the
Hi,
We are evaluating CEPH for deployment.
I was wondering if there are any current best practices around the
number of OSD's per node ?
e.g. We are looking at deploying 3 nodes, each with 72x SAS disks, and
2x 10gigabit Ethernet bonded.
Would this best be configured as 72 OSD's per
36 matches
Mail list logo
