Fio has a direct RBD engine which uses librbd. I've just had a quick look at
the code and I can't see an option in the latest fio to
specify datapool, but I'm not sure if librbd handles this all behind the
scenes. Might be worth a try.
From: ceph-users [mailto:ceph-users-boun...@lists.ceph.co
Fio with librbd takes pool as an argument. fio arguments looks like below.
ioengine=rbd
clientname=admin
pool=rbd
rbdname=test
If we can create the rbd image in a specific pool, we can run fio without any
issues. BTW i am using latest fio(built from github repo).
Varada
On Thursday 08 December
You can specify the -data-pool option while creating the rbd image.
Example:
rbd create rbdimg_EC1 --size 1024 --pool replicated_pool1 --data-pool ecpool
Once the image is created, you can add the image name(rdbimg_EC1) and the
replicated pool name(replicated_pool1) in the fio config file and set
Hi,
If you haven't already installed the previous branch, please try
wip-msgr-jewel-fix2 instead. That's a cleaner and more precise
solution to the real problem. :)
Any predictions when this fix will hit the Debian repositories?
Micha Krause
___
cep
On Thu, Dec 8, 2016 at 1:24 AM, Gregory Farnum wrote:
> Okay, we think we know what's happened; explanation at
> http://tracker.ceph.com/issues/18184 and first PR at
> https://github.com/ceph/ceph/pull/12372
>
> If you haven't already installed the previous branch, please try
> wip-msgr-jewel-fix2
On 16-12-08 09:32:02, Aravind Ramesh wrote:
> You can specify the -data-pool option while creating the rbd image.
> Example:
> rbd create rbdimg_EC1 --size 1024 --pool replicated_pool1 --data-pool ecpool
> Once the image is created, you can add the image name(rdbimg_EC1) and the
> replicated pool
The issue was due to a version mismatch in the rbd binary that I was using. It
was picking the old rbd instead of the ones I had compiled.
-Original Message-
From: Jason Dillaman [mailto:jdill...@redhat.com]
Sent: Thursday, December 08, 2016 2:05 AM
To: Aravind Ramesh
Cc: ceph-users@list
On 12/08/2016 11:24 AM, Ruben Kerkhof wrote:
> I've been running this on one of my servers now for half an hour, and
> it fixes the issue.
It's the same for me. ;)
~$ ceph -v
ceph version 10.2.4-1-g5d3c76c (5d3c76c1c6e991649f0beedb80e6823606176d9e)
Thanks for the help.
Bye.
Hello, folks,
I have problem with osd up after server reboot, all disk deployed with
ceph-deploy osd prepare --dmcrypt, i `ve tried ceph-deploy osd create
with no luck, same issue.
After server reboot osd dont come up, i need to run
*ceph-disk activate-lockbox /dev/sdb3 **
*
for example,
Hi David,
I'm surprised your message didn't get any echo yet. I guess it depends on how
many files your OSDs get to store on filesystem which depends essentialy on use
cases.
We're having similar issues with a 144 osd cluster running 2 pools. Each one
holds 100 M objects.One is replication x
Hi
I did a make install in my ceph build and also did make install on the fio and
ensured the latest binaries were installed. Now, fio is failing with below
errors for the rbd device with EC pool as data pool. I have shared the "rbd ls"
and my rbd.fio conf file details below. Let me know if you
Hey Khramchikhin,
On 12/08/2016 12:22 PM, Khramchikhin Nikolay wrote:
> Hello, folks,
>
> I have problem with osd up after server reboot, all disk deployed with
> ceph-deploy osd prepare --dmcrypt, i `ve tried ceph-deploy osd create
> with no luck, same issue.
>
> After server reboot osd dont
David, you might also be interested in the new Jewel 10.2.4 tool called
'ceph-objectstore-tool' from Josh.
It allows to split filestore directories offline
(http://tracker.ceph.com/issues/17220). Unfortunatly not merge apprently.
Regards,
Frédéric.
- Le 27 Sep 16, à 0:42, David Turner
On Thu, Dec 8, 2016 at 6:53 AM, Aravind Ramesh
wrote:
> I did a make install in my ceph build and also did make install on the fio
> and ensured the latest binaries were installed. Now, fio is failing with
> below errors for the rbd device with EC pool as data pool. I have shared the
> "rbd ls"
Hi Ceph Users !
I've got here a CEPH cluster: 6 nodes, 12 OSDs on HDD and SSD disks. All
journal OSDs on SSDs. 25 various HDDs in total.
We had several HDD failures in past, but every time - it was HDD failure
and it was never journal related. After replacing HDD, and recovery
procedures all was
No luck, i dont even have logs associate with ceph-disk at startup:
$journalctl -u ceph-
it shows me only standart ceph services, such ceph-osd@
but after ceph-disk activate-lockbox, i can get logs:
journalctl -u
ceph-disk@dev-disk-by\\x2dpartuuid-2c6532f7\\x2d4c5c\\x2d4fc7\\x2dab6c\\x2da7
On 16-12-08 08:45:44, Jason Dillaman wrote:
> On Thu, Dec 8, 2016 at 6:53 AM, Aravind Ramesh
> wrote:
> > I did a make install in my ceph build and also did make install on the fio
> > and ensured the latest binaries were installed. Now, fio is failing with
> > below errors for the rbd device wi
On 12/08/2016 03:09 PM, Khramchikhin Nikolay wrote:
> No luck, i dont even have logs associate with ceph-disk at startup:
In that case your udev rules probably didn't trigger. Have a look in the
logs and see if udev invokes (/usr/sbin/ceph-disk --log-stdout -v
trigger /dev/$name)
you could als
Hi,
I have a CephFS cluster that is currently unable to start the mds server as
it is hitting an assert, the extract from the mds log is below, any
pointers are welcome:
ceph version 10.2.3 (ecc23778eb545d8dd55e2e4735b53cc93f92e65b)
2016-12-08 14:50:18.577038 7f7d9faa3700 1 mds.0.47077 handle_m
Hello,
I have a similar problem. I am not sure it's the same, but if i can help.
I'm in jewel ( upgrade from firefly ) on jessie.
The temporary solution that i found to start the OSD is to force udev to
launch its triggers :
udevadm trigger --action=add
Regards
--
--
On Thu, Dec 8, 2016 at 3:11 PM, Sean Redmond wrote:
> Hi,
>
> I have a CephFS cluster that is currently unable to start the mds server as
> it is hitting an assert, the extract from the mds log is below, any pointers
> are welcome:
>
> ceph version 10.2.3 (ecc23778eb545d8dd55e2e4735b53cc93f92e65b)
Hi,
We had no changes going on with the ceph pools or ceph servers at the time.
We have however been hitting this in the last week and it maybe related:
http://tracker.ceph.com/issues/17177
Thanks
On Thu, Dec 8, 2016 at 3:34 PM, John Spray wrote:
> On Thu, Dec 8, 2016 at 3:11 PM, Sean Redmon
Udev doesnt trigger at start. But i have correct rules at udev-rules file.
OS - debian jessie.ceph - jewel.
Hello,
I have a similar problem. I am not sure it's the same, but if i can help.
I'm in jewel ( upgrade from firefly ) on jessie.
The temporary solution that i foun
Thanks Jason,
Yes, after enabling the flag, it is able to run IOs on it.
Aravind
-Original Message-
From: Jason Dillaman [mailto:jdill...@redhat.com]
Sent: Thursday, December 08, 2016 7:16 PM
To: Aravind Ramesh
Cc: Venky Shankar ; ceph-users@lists.ceph.com;
n...@fisk.me.uk
Subject: Re:
[..]
>
>
> Yep, it`s same problem. It looks like bug. Manual udevadm trigger
> worked fine.
Good, seems like at least the ceph part is functioning correctly.
>
[..]
___
ceph-users mailing list
ceph-users@lists.ceph.com
http://lists.ceph.com/listinfo
Hey Dan,
I had the same issue that Jacek had after changing my OS and Ceph version
from Ubuntu 14 - Hammer to Centos 7 - Jewel. I was also able to recover
from the failure by renaming the .ldb files to .sst files.
Do you know why this works?
Is it just because leveldb changed the file naming st
Hi
I am working alongside Sean with this assertion issue. We see problems with
the stray calls when the MDS starts, the last entry in the log before the
assertion failure is a reference: try_remove_dentries_for_stray. I have
provided a link for the ceph-collect logs, which we're collected after
sp
I don't want to retype it all, but you guys might be interested in the
discussion under section 3 of this post here:
http://lists.ceph.com/pipermail/ceph-users-ceph.com/2016-September/012987.html
basically the gist of it is:
1) Make sure SELinux isn't doing security xattr lookups for link/unli
On Thu, Dec 8, 2016 at 3:45 PM, Sean Redmond wrote:
> Hi,
>
> We had no changes going on with the ceph pools or ceph servers at the time.
>
> We have however been hitting this in the last week and it maybe related:
>
> http://tracker.ceph.com/issues/17177
Oh, okay -- so you've got corruption in y
Hi John.
I have been hitting that issue also although have not seen any asserts in my
mds yet.
Could you please clarify a bit further your proposal about manually removing
omap info from strays? Should it be applied:
- to the problematic replicas of the stray object which triggered the
inconsi
Hi John,
Thanks for your pointers, I have extracted the onmap_keys and onmap_values
for an object I found in the metadata pool called '600.' and
dropped them at the below location
https://www.dropbox.com/sh/wg6irrjg7kie95p/AABk38IB4PXsn2yINpNa9Js5a?dl=0
Could you explain how is it possib
Hi John / All
Thank you for the help so far.
To add a further point to Sean's previous email, I see this log entry
before the assertion failure:
-6> 2016-12-08 15:47:08.483700 7fb133dca700 12
mds.0.cache.dir(1000a453344) remove_dentry [dentry
#100/stray9/1000a453344/config [2,head] auth NULL
"Hello List,
This could be transformed a bugreport if anyone feels like, I just
kind of share my harsh experience of today.
We had a few OSDs getting full while others being below 40%; while
these were properly weighted (full ones of 800GB being 0.800 and
fairly empty ones being 2.7TB and 2.700 w
I've seen this before and would recommend upgrading from Hammer.
On 12/08/2016 04:26 PM, Peter Gervai wrote:
"Hello List,
This could be transformed a bugreport if anyone feels like, I just
kind of share my harsh experience of today.
We had a few OSDs getting full while others being below 40%;
On Wed, Dec 7, 2016 at 5:45 PM, Andreas Gerstmayr
wrote:
> Hi,
>
> does the CephFS kernel module (as of kernel version 4.8.8) support
> parallel reads of file stripes?
> When an application requests a 500MB block from a file (which is
> splitted into multiple objects and stripes on different OSDs)
Hi,
so - we're considering moving omap out to faster media than our rather slow
spinning rust. There's been some discussion around this here:
https://github.com/ceph/ceph/pull/6421
Since this hasn't landed in jewel, or the ceph-disk convenience bits -
we're thinking of "other ways" of doing this.
Thanks for your response!
2016-12-09 1:27 GMT+01:00 Gregory Farnum :
> On Wed, Dec 7, 2016 at 5:45 PM, Andreas Gerstmayr
> wrote:
>> Hi,
>>
>> does the CephFS kernel module (as of kernel version 4.8.8) support
>> parallel reads of file stripes?
>> When an application requests a 500MB block from a
Confused ...
a few OSDs down and cluster done the recovery and reblanced to HEALTH OK
state.
Now I can could that down OSDs are down state from crushmap and are not
part of OSD up or in state.
After 5 days or says, still the same state.
How or when Ceph will make the down state OSDs to out state? I
On Fri, Dec 9, 2016 at 3:28 PM, M Ranga Swami Reddy
wrote:
> Confused ...
> a few OSDs down and cluster done the recovery and reblanced to HEALTH OK
> state.
> Now I can could that down OSDs are down state from crushmap and are not
> part of OSD up or in state.
> After 5 days or says, still the s
no noout and other flags set...
How do we confirm if the down OSD is down of the cluster?
Thanks
Swami
On Fri, Dec 9, 2016 at 11:18 AM, Brad Hubbard wrote:
>
>
> On Fri, Dec 9, 2016 at 3:28 PM, M Ranga Swami Reddy
> wrote:
>
>> Confused ...
>> a few OSDs down and cluster done the recovery an
Hi,
We have managed to work around this problem.
It seems that one of the objects lost due to the SSD failure was part of
the hitset for the cache tier. The OSD was dying after an attempt to open
and read the object's data file.
In an effort to progress I have found out which file is missing by
41 matches
Mail list logo
