Hi guys,
___
ceph-users mailing list
ceph-users@lists.ceph.com
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
2014-10-21 7:40 GMT+08:00 Lionel Bouton :
> Hi,
>
> Le 21/10/2014 01:10, 池信泽 a écrit :
>
> Thanks.
>
>Another reason is the checksum in the attr of object used for deep scrub
> in EC pools should be computed when modify the object. When supporting the
> random write, We should caculate the whol
2014-10-20 22:39 GMT+08:00 Wido den Hollander :
> On 10/20/2014 03:25 PM, 池信泽 wrote:
>> hi, cephers:
>>
>> When I look into the ceph source code, I found the erasure code pool
>> not support
>> the random write, it only support the append write. Why? Is that random
>> write of is erasure code
Hi all,
thank you for your support, now the file system is not degraded any more. Now I
have a minus degrading :-)
2014-10-21 10:15:22.303139 mon.0 [INF] pgmap v43376478: 3328 pgs: 3281
active+clean, 47 active+remapped; 1609 GB data, 5022 GB used, 1155 GB / 6178 GB
avail; 8034B/s rd, 3548KB/s
Dear Users,
whenever I run the command
ceph-deploy
with whatever option I got at the end this error message:
Error in sys.exitfunc:
The command seems to work.
Can it be ignored?
Anyone else had this problem?
I'm running a Centos 6.5
Thank you.
Cheer
Luca Mazzaferro
<>_
Hello cephers,
Does anyone know when is the planned release date for Giant?
Cheers
Andrei
___
ceph-users mailing list
ceph-users@lists.ceph.com
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Hi Dan and Ashish,
Many thanks for the replies.
It is partly for the learning( and perhaps a production system in future) that
we are playing with ceph.
I guess, to start with, a simple set up of ,may be 4 vms, where one vm is admin
node and one is mons and 2 are osds ( as suggested in pre-fl
Le 21/10/2014 09:31, Nicheal a écrit :
> 2014-10-21 7:40 GMT+08:00 Lionel Bouton :
>> Hi,
>>
>> Le 21/10/2014 01:10, 池信泽 a écrit :
>>
>> Thanks.
>>
>>Another reason is the checksum in the attr of object used for deep scrub
>> in EC pools should be computed when modify the object. When supportin
Hello,
On Tue, 21 Oct 2014 10:53:54 +0100 Ranju Upadhyay wrote:
> Hi Dan and Ashish,
>
> Many thanks for the replies.
>
> It is partly for the learning( and perhaps a production system in
> future) that we are playing with ceph.
>
Learning about Ceph (configuration, operation wise) with VMs is
Agreed. The "getting started" instructions walk you through creating most of a
cluster then adding once more node piecemeal, but this seems needlessly complex
for a beginner. I found the easiest way to get up and running was with three
VMs with one or two logical partitions for OSDs on each and
On Tue, 21 Oct 2014, Andrei Mikhailovsky wrote:
> Hello cephers,
>
> Does anyone know when is the planned release date for Giant?
Looking at another week or two. There are still a few outstanding issues
we'd like to squash!
sage
___
ceph-users mailin
Hi cephers,
We have two pgs that are stuck in 'incomplete' state across two different
pools:
pg 2.525 is stuck inactive since forever, current state incomplete, last acting
[55,89]
pg 0.527 is stuck inactive since forever, current state incomplete, last acting
[55,89]
pg 0.527 is stuck unclean
Thanks, Sage,
Can't wait to try it out and see if there are any improvements in the caching
pool tier.
Cheers
Andrei
- Original Message -
> From: "Sage Weil"
> To: "Andrei Mikhailovsky"
> Cc: "ceph-users"
> Sent: Tuesday, 21 October, 2014 2:27:55 PM
> Subject: Re: [ceph-users]
Hi Craig,
> It's part of the way the CRUSH hashing works. Any change to the CRUSH map
> causes the algorithm to change slightly.
Dan@cern could not replicate my observations, so I plan to follow his
procedure (fake create an OSD, wait for rebalance, remove fake OSD) in the
near future to see i
That will fix itself over time. remapped just means that Ceph is moving
the data around. It's normal to see PGs in the remapped and/or backfilling
state after OSD restarts.
They should go down steadily over time. How long depends on how much data
is in the PGs, how fast your hardware is, how ma
>>Thank you for your quick response! Okay I see, is there any preferred
>>clustered FS in this case? OCFS2, GFS?
Hi, I'm using ocfs2 on top of rbd in production, works fine. (you need to
disable writeback/rbd_cache)
- Mail original -
De: "Mihály Árva-Tóth"
À: "Sean Redmond"
Cc: c
On Tuesday, October 21, 2014, Chad Seys wrote:
> Hi Craig,
>
> > It's part of the way the CRUSH hashing works. Any change to the CRUSH
> map
> > causes the algorithm to change slightly.
>
> Dan@cern could not replicate my observations, so I plan to follow his
> procedure (fake create an OSD, wai
I'm still pretty new at Ceph so take this with a grain of salt.
1. In our experience, we have tried SSD journals and bcache, we have had
more stability and performance by just using SSD journals. We have created
an SSD pool with the rest of the space and it did not perform much better
A small update on this, I rebooted all of the Ceph nodes and was able to then
query one of the misbehaving pgs.
I've attached the query for pg 2.525.
{\rtf1\ansi\ansicpg1252\cocoartf1138\cocoasubrtf510
{\fonttbl\f0\fswiss\fcharset0 Helvetica;}
{\colortbl;\red255\green255\blue255;}
\margl1440\ma
Whoops, didn't mean to attach the file as rtf. Plaintext attached
{ "state": "incomplete",
"epoch": 224352,
"up": [
55,
89],
"acting": [
55,
89],
"info": { "pgid": "2.525",
"last_update": "0'0",
"last_complete": "0'0",
"log_tail": "0'0",
Hi,
I've yet to install 0.80.7 on one node to confirm its stability and use
the new IO prirority tuning parameters enabling prioritized access to
data from client requests.
In the meantime, faced with large slowdowns caused by resync or external
IO load (although external IO load is not expected
I configured a three-monitor Ceph cluster following the manual
instructions at
http://ceph.com/docs/v0.80.5/install/manual-deployment/ and
http://ceph.com/docs/master/rados/operations/add-or-rm-mons/ . The
monitor cluster came up without a problem, and seems to be fine. "ceph
-s" currently shows t
Hi
I just built a new cluster using this quickstart instructions:
http://ceph.com/docs/master/start/
And here is what I am seeing:
# time for i in {1..10}; do echo $i > $i.txt ; done
real 0m0.081s
user 0m0.000s
sys 0m0.004s
And if I try to repeat the same command (when files already created):
On Tue, Oct 21, 2014 at 10:15 AM, Lionel Bouton wrote:
> Hi,
>
> I've yet to install 0.80.7 on one node to confirm its stability and use
> the new IO prirority tuning parameters enabling prioritized access to
> data from client requests.
>
> In the meantime, faced with large slowdowns caused by re
Are these tests conducted using a local fs on RBD, or using CephFS?
If CephFS, do you have multiple clients mounting the FS, and what are
they doing? What client (kernel or ceph-fuse)?
-Greg
Software Engineer #42 @ http://inktank.com | http://ceph.com
On Tue, Oct 21, 2014 at 9:05 AM, Sergey Nazar
It is CephFS mounted via ceph-fuse.
I am getting the same results not depending on how many other clients
are having this fs mounted and their activity.
Cluster is working on Debian Wheezy, kernel 3.2.0-4-amd64.
On Tue, Oct 21, 2014 at 1:44 PM, Gregory Farnum wrote:
> Are these tests conducted us
Hi Gregory,
Le 21/10/2014 19:39, Gregory Farnum a écrit :
> On Tue, Oct 21, 2014 at 10:15 AM, Lionel Bouton
> wrote:
>> [...]
>> Any thought? Is it based on wrong assumptions? Would it prove to be a
>> can of worms if someone tried to implement it?
> Yeah, there's one big thing you're missing: w
Can you enable debugging on the client ("debug ms = 1", "debug client
= 20") and mds ("debug ms = 1", "debug mds = 20"), run this test
again, and post them somewhere for me to look at?
While you're at it, can you try rados bench and see what sort of
results you get?
-Greg
Software Engineer #42 @ h
On 10/21/2014 01:06 PM, Lionel Bouton wrote:
Hi Gregory,
Le 21/10/2014 19:39, Gregory Farnum a écrit :
On Tue, Oct 21, 2014 at 10:15 AM, Lionel Bouton wrote:
[...]
Any thought? Is it based on wrong assumptions? Would it prove to be a
can of worms if someone tried to implement it?
Yeah, there
After more than 10 hours the same situation, I don’t think it will fix self
over time. How I can find out what is the problem.
Am 21.10.2014 um 17:28 schrieb Craig Lewis
mailto:cle...@centraldesktop.com>>:
That will fix itself over time. remapped just means that Ceph is moving the
data aroun
I've had issues magically fix themselves over night after waiting/trying
things for hours.
On Tue, Oct 21, 2014 at 1:02 PM, Harald Rößler
wrote:
> After more than 10 hours the same situation, I don’t think it will fix
> self over time. How I can find out what is the problem.
>
>
> Am 21.10.2014
I enabled logging and performed same tests.
Here is the link on archive with logs, they are only from one node
(from the node where active MDS was sitting):
https://www.dropbox.com/s/80axovtoofesx5e/logs.tar.gz?dl=0
Rados bench results:
# rados bench -p test 10 write
Maintaining 16 concurrent wr
Ouch, I think client log is missing.
Here it goes:
https://www.dropbox.com/s/650mjim2ldusr66/ceph-client.admin.log.gz?dl=0
On Tue, Oct 21, 2014 at 3:22 PM, Sergey Nazarov wrote:
> I enabled logging and performed same tests.
> Here is the link on archive with logs, they are only from one node
> (f
In that case, take a look at ceph pg dump | grep remapped. In the up or
active column, there should be one or two common OSDs between the stuck PGs.
Try restarting those OSD daemons. I've had a few OSDs get stuck scheduling
recovery, particularly around toofull situations.
I've also had Robert'
Hi Sage,
Yes, I know about rbd diff, but the motivation behind this was to be able to
dump an entire RBD pool via RADOS to another cluster, as our primary cluster
uses quite expensive SSD storage and we would like to avoid constantly keeping
one snapshot for every RBD image.
The idea would b
On Mon, 20 Oct 2014 11:07:43 +0200 Leszek Master wrote:
> 1) If i want to use cache tier should i use it with ssd journaling or i
> can get better perfomance using more ssd GB for cache tier?
>
>From reading what others on this ML experienced and what Robert already
pointed out, cache tiering is
Hello,
I'm trying to change the value of mon_osd_down_out_subtree_limit from rack
to something, anything else with ceph 0.80.(6|7).
Using injectargs it tells me that this isn't a runtime supported change
and changing it in the config file (global section) to either host or room
has no effect.
C
I've gotten myself into the position of having ~100 incomplete PGs. All of my
OSDs are up+in (and I've restarted them all one by one).
I was in the process of rebalancing after altering my CRUSH map when I lost an
OSD backing disk. I replaced that OSD and it seemed to be backfilling well.
Durin
38 matches
Mail list logo
