On Mon, Jun 20, 2016 at 5:28 AM, Daleep Singh Bais wrote:
> Dear All,
>
> I am getting below error message while trying to build Jewel on ARM. Any
> help / suggestion will be appreciated.
>
> g++: error: unrecognized command line option '-momit-leaf-frame-pointer'
> g++: error: unrecognized comman
On Mon, Jun 20, 2016 at 8:33 AM, Daniel Swarbrick
wrote:
> We have just updated our third cluster from Infernalis to Jewel, and are
> experiencing similar issues.
>
> We run a number of KVM virtual machines (qemu 2.5) with RBD images, and
> have seen a lot of D-state processes and even jbd/2 timeo
On Wednesday, June 22, 2016, Kenneth Waegeman
wrote:
> Hi all,
>
> In Jewel ceph fs snapshots are still experimental. Does someone has a clue
> when this would become stable, or how experimental this is ?
>
We're not sure yet. Probably it will follow stable multi-MDS; we're
thinking about redoin
. Are
> any of the know issues with snapshots issues that result in the loss of
> non-snapshot data or a cluster?
> On Jun 22, 2016 2:16 PM, "Gregory Farnum" wrote:
>
>> On Wednesday, June 22, 2016, Kenneth Waegeman
>> wrote:
>>
>>> Hi all,
>
On Sat, Jun 25, 2016 at 11:22 AM, Mike Miller wrote:
> Hi,
>
> what is the meaning of the directory "current.remove.me.846930886" is
> /var/lib/ceph/osd/ceph-14?
If you're using btrfs, I believe that's a no-longer-required snapshot
of the current state of the system. If you're not, I've no idea w
On Thu, Jun 30, 2016 at 9:09 AM, Mauricio Garavaglia
wrote:
> Hello,
>
> What's the expected behavior of a host that has a cephfs mounted and is then
> blacklisted? It doesn't seem to fail in a consistent way. Thanks
Well, once blacklisted it won't be allowed to make updates to the OSDs
or MDS. D
On Wed, Jun 29, 2016 at 10:50 PM, Goncalo Borges
wrote:
> Hi Shinobu
>
>> Sorry probably I don't understand your question properly.
>> Is what you're worry about that object mapped to specific pg could be
>> overwritten on different osds?
>
> Not really. I was worried by seeing object sizes chang
On Wed, Jun 29, 2016 at 2:02 PM, Daniel Davidson
wrote:
> I am starting to work with and benchmark our ceph cluster. While throughput
> is so far looking good, metadata performance so far looks to be suffering.
> Is there anything that can be done to speed up the response time of looking
> throug
On Thu, Jun 30, 2016 at 1:03 PM, Dzianis Kahanovich wrote:
> Upgraded infernalis->jewel (git, Gentoo). Upgrade passed over global
> stop/restart everything oneshot.
>
> Infernalis: e5165: 1/1/1 up {0=c=up:active}, 1 up:standby-replay, 1 up:standby
>
> Now after upgrade start and next mon restart,
Thanks for the report; created a ticket and somebody will get on it
shortly. http://tracker.ceph.com/issues/16592
-Greg
On Sun, Jul 3, 2016 at 5:55 PM, Bill Sharer wrote:
> I was working on a rolling upgrade on Gentoo to Jewel 10.2.2 from 10.2.0.
> However now I can't get a monitor quorum going
On Mon, Jul 4, 2016 at 12:38 PM, Dzianis Kahanovich wrote:
> Gregory Farnum пишет:
>> On Thu, Jun 30, 2016 at 1:03 PM, Dzianis Kahanovich wrote:
>>> Upgraded infernalis->jewel (git, Gentoo). Upgrade passed over global
>>> stop/restart everything oneshot.
>>&g
On Thu, Jul 7, 2016 at 1:07 AM, 秀才 wrote:
> Hi,All:)
>
> i have made a cache-tier,
> but i do not know message 'removed_snaps
> [1~1,3~6,b~6,13~c,21~4,26~1,28~1a,4e~4,53~5,5c~5,63~1,65~4,6b~4]'.
> i have not snapped any thing yet.
When you take snapshots, it generally creates a lot of tracking da
I'm not looking at the docs, but I think you need an "emit" statement after
every choose.
-Greg
On Monday, July 11, 2016, George Shuklin wrote:
> Hello.
>
> I want to try CRUSH rule with following idea:
> take one OSD from root with SSD drives (and use it as primary).
> take two OSD from root wi
Oh, is this one of your custom-built packages? Are they using
tcmalloc? That difference between VSZ and RSS looks like a glibc
malloc problem.
-Greg
On Mon, Jul 11, 2016 at 12:04 AM, Goncalo Borges
wrote:
> Hi John...
>
> Thank you for replying.
>
> Here is the result of the tests you asked but I
uleset 1
> type replicated
> min_size 1
> max_size 4
> step take ssd
> step choose firstn 2 type room
> step chooseleaf firstn 2 type host
> step emit
> }
>
> Bob
>
> On Mon, Jul 11, 2016 at 9:19 AM, Gregory Farnum wrote:
>>
>> I'm not looking at the docs,
I'm not at all sure that rados cppool actually captures everything (it
might). Doug has been working on some similar stuff for disaster
recovery testing and can probably walk you through moving over.
But just how large *is* your metadata pool in relation to others?
Having a too-large pool doesn't
On Thu, Jul 14, 2016 at 1:42 AM, txm wrote:
> I am a user of cephfs.
>
> Recently i met a problem by using the cephfs-journal-tool.
>
> There were some strange things happened below.
>
> 1.After use the cephfs-journal-tool and cephfs-table-tool(i came up with the
> "negative object nums” issues,
Specifically, this has caused trouble with certain (common?) rsync
configurations.
-Greg
On Monday, July 18, 2016, David wrote:
> Thanks Zheng, I should have checked that.
>
> Sean, from the commit:
>
> When rbytes mount option is enabled, directory size is recursive size.
> Recursive size is no
I'm not familiar with how it's set up but skimming and searching
through the code I'm not seeing anything, no. We've got a chown but no
chmod. That's a reasonably feature idea though, and presumably you
could add a chmod it to your init scripts?
-Greg
On Mon, Jul 18, 2016 at 3:02 AM, Stefan Priebe
On Tue, Jul 19, 2016 at 9:39 AM, Patrick Donnelly wrote:
> On Tue, Jul 19, 2016 at 10:25 AM, Fabiano de O. Lucchese
> wrote:
>> I configured the cluster to replicate data twice (3 copies), so these
>> numbers fall within my expectations. So far so good, but here's comes the
>> issue: I configured
n in write throughput to around 170 Mbps. The system
> took about 5-6 seconds before it started writing any files and was
> constantly at 150 – 180 Mbps write through put when the directory was
> mounted using FUSE.
>
> Any additional thoughts? Would the problem be due to my NFS clien
On Thu, Jul 21, 2016 at 11:54 AM, Salwasser, Zac wrote:
> Rephrasing for brevity – I have a monitor store that is 69GB and won’t
> compact any further on restart or with ‘tell compact’. Has anyone dealt
> with this before?
The monitor can't trim OSD maps over a period where PGs are unclean;
you'
All,
I spent several days last week examining our current snapshot
implementation and thinking about how it could be improved. As part of
that ongoing effort, I'd love to know what user expectations are about
behavior.
(I'm going to open up a ceph-devel thread on the implementation
details shortly,
On Mon, Jul 25, 2016 at 7:38 PM, Oliver Dzombic wrote:
> Hi,
>
> currently some productive stuff is down, because it can not be accessed
> through cephfs.
>
> Client server restart, did not help.
> Cluster restart, did not help.
>
> Only ONE directory inside cephfs has this issue.
>
> All other di
4M 85.10 1.72 288
> 11 0.21799 1.0 223G 193G 31155M 86.38 1.75 285
> TOTAL 53926G 26654G 27272G 49.43
> MIN/MAX VAR: 0.72/1.84 STDDEV: 21.46
>
>
> --
> Mit freundlichen Gruessen / Best regards
>
> Oliver Dzombic
> IP-Interactive
>
> mailt
On Tue, Jul 26, 2016 at 1:53 PM, Mike Miller wrote:
> Hi,
>
> we have started to migrate user homes to cephfs with the mds server 32GB
> RAM. With multiple rsync threads copying this seems to be undersized; the
> mds process consumes all memory 32GB fitting about 4 million caps.
>
> Any hardware r
On Wed, Jul 27, 2016 at 9:17 AM, Andrey Ptashnik wrote:
> Hello team,
>
> We are looking for ways to store metadata with objects and make this metadata
> searchable.
> For example if we store an image of the car in Ceph we would like to be able
> to attach metadata like model, make, year, damage
On Wed, Jul 27, 2016 at 2:51 PM, Patrick Donnelly wrote:
> On Mon, Jul 25, 2016 at 5:41 PM, Gregory Farnum wrote:
>> Some specific questions:
>> * Right now, we allow users to rename snapshots. (This is newish, so
>> you may not be aware of it if you've been using sna
On Wed, Jul 27, 2016 at 6:13 PM, Goncalo Borges
wrote:
> Dear cephfsers :-)
>
> We saw some weirdness in cephfs that we do not understand.
>
> We were helping some user which complained that her batch system job outputs
> were not produced in cephfs.
>
> Please note that we are using ceph-fuse (je
On Wed, Jul 27, 2016 at 6:37 PM, Goncalo Borges
wrote:
> Hi Greg
>
> Thanks for replying. Answer inline.
>
>
>
>>> Dear cephfsers :-)
>>>
>>> We saw some weirdness in cephfs that we do not understand.
>>>
>>> We were helping some user which complained that her batch system job
>>> outputs
>>> were
On Thu, Jul 28, 2016 at 2:44 PM, Alexandre Oliva wrote:
> On Jul 25, 2016, Gregory Farnum wrote:
>
>> * Right now, we allow users to rename snapshots. (This is newish, so
>> you may not be aware of it if you've been using snapshots for a
>> while.) Is that an importa
On Sun, Jul 31, 2016 at 6:53 PM, Goncalo Borges
wrote:
> Dear cephers.
>
> I would like to request some clarification on migrating from legacy to
> optimal (jewel) tunables.
>
> We have recently migrated from infernalis to Jewel. However, we are still
> using legacy tunables.
>
> All our ceph infr
On Wed, Aug 3, 2016 at 3:50 PM, Somnath Roy wrote:
> Probably, it is better to move to latest master and reproduce this defect.
> Lot of stuff has changed between this.
> This is a good test case and I doubt any of us testing by enabling fsck() on
> mount/unmount.
Given that the allocator keeps
On Thu, Aug 4, 2016 at 8:57 PM, Goncalo Borges
wrote:
> Dear cephers...
>
> I am looking for some advice on migrating from legacy tunables to Jewel
> tunables.
>
> What would be the best strategy?
>
> 1) A step by step approach?
> - starting with the transition from bobtail to firefly (and, in
nk we should set?
>
> What do you reckon?
That is generally the big one, but I think you'll need advice from
people who actually run clusters to see if there's anything more
that's useful. :)
-Greg
>
> Cheers
>
> Goncalo
>
>
> On 08/09/2016 09:26 AM, Gre
On Tue, Aug 9, 2016 at 2:00 AM, Kenneth Waegeman
wrote:
> Hi,
>
> I did a diff on the directories of all three the osds, no difference .. So I
> don't know what's wrong.
omap (as implied by the omap_digest complaint) is stored in the OSD
leveldb, not in the data directories, so you wouldn't expec
On Sunday, February 24, 2013 at 1:48 PM, Andrey Korolyov wrote:
> Hi,
>
> "filestore_queue_max_ops": "50",
> "filestore_flusher": "true",
>
> Just checked that this options are set with values above instead of
> config` 500 and 0 as mentioned in docs, and second option has an
> observable perfo
On Sunday, February 24, 2013 at 2:42 PM, Andrey Korolyov wrote:
> On Mon, Feb 25, 2013 at 2:37 AM, Gregory Farnum (mailto:g...@inktank.com)> wrote:
> > On Sunday, February 24, 2013 at 1:48 PM, Andrey Korolyov wrote:
> > > Hi,
> > >
> > > "filestore_qu
On Sunday, February 24, 2013 at 2:59 PM, Andrey Korolyov wrote:
> On Mon, Feb 25, 2013 at 2:49 AM, Gregory Farnum (mailto:g...@inktank.com)> wrote:
> > On Sunday, February 24, 2013 at 2:42 PM, Andrey Korolyov wrote:
> > > On Mon, Feb 25, 2013 at 2:37 AM, Gregory
On Sun, Feb 24, 2013 at 3:10 PM, Andrey Korolyov wrote:
> On Mon, Feb 25, 2013 at 3:07 AM, Gregory Farnum wrote:
>> Try
>> ceph --keyfile admin osd tell \* injectargs "--filestore-flusher=0"
>> (add the equals sign) and let me know if that works. Looks like may
On Mon, Feb 25, 2013 at 11:14 AM, Gregory Farnum wrote:
> On Sun, Feb 24, 2013 at 3:10 PM, Andrey Korolyov wrote:
>> On Mon, Feb 25, 2013 at 3:07 AM, Gregory Farnum wrote:
>>> Try
>>> ceph --keyfile admin osd tell \* injectargs "--filestore-flusher=0"
>
On Mon, Feb 25, 2013 at 8:44 AM, Sage Weil wrote:
> On Mon, 25 Feb 2013, Steffen Thorhauer wrote:
>> Hi,
>> I've found out, what I make wrong: stop the cluster and forget a client,
>> which as mounting the cephfs. I simply forget the client.
>> With a
>> ceph mds newfs 0 1 --yes-i-really-mean-it
On Wed, Feb 27, 2013 at 3:20 AM, Yann ROBIN wrote:
> Hi,
>
> What could go wrong with this command ? If something goes wrong, is there any
> chance of recovery ?
> As this is a production cluster, I don't want to test it if there is a chance
> of losing my cluster an data.
It could go horribly
I haven't checked the logs at all, but my first question is what the
memory utilization on that node looked like during that time, and if
there are any dmesg warnings.
(Other than that I should leave this to probably Alex and Sam.)
-Greg
On Thu, Feb 28, 2013 at 1:44 PM, Nick Bartos wrote:
> When
On Fri, Mar 1, 2013 at 8:17 AM, Scott Kinder wrote:
> In my ceph.conf file, I set the options under the [osd] section:
>
> osd pool default pg num = 133
> osd pool default pgp num = 133
You'll want to place those in the global section rather than the OSD
section — the pools are generated on the m
This doesn't cover reads from the same objects though, and in fact
Ceph deliberately sends all object reads to the primary copy. There
are various options (which are variously implemented) to turn this off
but turning it into a proper striped reads system or something would
take some work.
-Greg
O
On Fri, Mar 1, 2013 at 1:53 PM, Karsten Becker
wrote:
> Hi,
>
> I'm new to Ceph. I currently find no answer in the official docs for the
> following question.
>
> Can Ceph filesystems be used concurrently by clients, both when
> accessing via RBD and CephFS? Concurrently means in terms of multiple
This doesn't sound quite right but I'm not sure if the problem is a
terminology disconnect or a concept one. Let's go through them
On Fri, Mar 1, 2013 at 3:08 PM, McNamara, Bradley
wrote:
> I'm new, too, and I guess I just need a little clarification on Greg's
> statement. The RBD filesystem
>
On Mon, Mar 4, 2013 at 12:19 AM, Olivier Bonvalet wrote:
> Hello,
>
> I have a running cluster, which use the (previous) default CRUSH rule,
> with :
> step take default
> step choose firstn 0 type osd
> step emit
>
> Since I have multiple OSD in the same host, I need to ch
On Wednesday, March 6, 2013 at 2:43 AM, Félix Ortega Hortigüela wrote:
> Hi
> I'm running ceph v56.3 over debian-wheezy, with the ceph.com
> (http://ceph.com) debs.
> My setup is three servers with 6 disk each. I have 5 disks on each server
> dedicated to osd's and the disk left is dedicated to t
On Monday, March 11, 2013, Igor Laskovy wrote:
> Hi there!
>
> I have Ceph FS cluster version 0.56.3. This is 3 nodes with XFS on disks
> and with minimum options in ceph.conf in my lab and I do some crush
> testing.
> One of the of several tests is lost connect to replication network only.
> What
On Sunday, March 17, 2013 at 4:46 AM, Andrey Korolyov wrote:
> Hi,
>
> from osd tree:
>
> -16 4.95 host 10.5.0.52
> 32 1.9 osd.32 up 2
> 33 1.05 osd.33 up 1
> 34 1 osd.34 up 1
> 35 1 osd.35 up 1
>
> df -h:
> /dev/sdd3 3.7T 595G 3.1T 16% /var/lib/ceph/osd/32
> /dev/sde3 3.7T 332G 3.4T 9% /var/
On Sunday, March 17, 2013 at 9:09 AM, Andrey Korolyov wrote:
> On Sun, Mar 17, 2013 at 7:56 PM, Gregory Farnum (mailto:g...@inktank.com)> wrote:
> > On Sunday, March 17, 2013 at 4:46 AM, Andrey Korolyov wrote:
> > > Hi,
> > >
> > > from osd tree:
> &g
On Friday, March 15, 2013 at 2:02 AM, Li, Chen wrote:
> I need to create the directory “/var/lib/ceph/mds/mds.$id ”by hand, right ?
>
> I start the service as you said, and it is succeed.
> But, no “mds.$id” directory exist.
> Will this affect it working?
>
> And, what will be installed in the
RADOS doesn't store a list of objects. The RADOS Gateway uses a separate data
format on top of objects stored in RADOS, and it keeps a per-user list of
buckets and a per-bucket index of objects as "omap" objects in the OSDs (which
ultimately end up in a leveldb store). A bucket index is currentl
On Sunday, March 17, 2013 at 9:25 AM, Andrey Korolyov wrote:
> On Sun, Mar 17, 2013 at 8:14 PM, Gregory Farnum (mailto:g...@inktank.com)> wrote:
> > On Sunday, March 17, 2013 at 9:09 AM, Andrey Korolyov wrote:
> > > On Sun, Mar 17, 2013 at 7:56 PM, Gregory Farnum > &g
On Sunday, March 17, 2013 at 4:03 PM, Mark Nelson wrote:
> On 03/17/2013 05:40 PM, Matthieu Patou wrote:
> > Hello all,
> >
> > Our dev environment are quite I/O intensive but didn't require much
> > space (~20G per dev environment), for the moment our dev machines are
> > served by VMWare and th
On Sunday, March 17, 2013 at 5:49 PM, Yehuda Sadeh wrote:
> Not at the moment. We had some discussions about "blind" buckets, it's
> definitely on our mind, but we're not there yet.
>
> Yehuda
>
> On Sun, Mar 17, 2013 at 3:09 PM, Rustam Aliyev (mailto:rustam.li...@code.az)> wrote:
> > Thanks for
On Wed, Mar 20, 2013 at 5:06 PM, Darryl Bond wrote:
> I have a cluster of 3 hosts each with 2 SSD and 4 Spinning disks.
> I used the example in th ecrush map doco to create a crush map to place
> the primary on the SSD and replica on spinning disk.
>
> If I use the example, I end up with objects r
This should actually be "down", not "out". Out will trigger a data migration. :)
-Greg
Software Engineer #42 @ http://inktank.com | http://ceph.com
On Wed, Mar 27, 2013 at 8:36 AM, John Nielsen wrote:
> Sébastien Han actually covers this on his blog:
>
> http://www.sebastien-han.fr/blog/2012/08
This is the perfectly normal distinction between "down" and "out". The
OSD has been marked down but there's a timeout period (default: 5
minutes) before it's marked "out" and the data gets reshuffled (to
avoid starting replication on a simple reboot, for instance).
-Greg
Software Engineer #42 @ htt
168.195.33:6789/0,c=192.168.195.35:6789/0},
> election epoch 6, quorum 0,1,2 a,b,c
>osdmap e28: 24 osds: 23 up, 24 in
> pgmap v454: 4800 pgs: 4422 active+clean, 378 active+degraded; 1800
> MB data, 3780 MB used, 174 TB / 174 TB avail; 39/904 degraded (4.314%)
>mdsmap e1
default
> step chooseleaf firstn 0 type rack
> step emit
> }
> rule metadata {
> ruleset 1
> type replicated
> min_size 1
> max_size 10
> step take default
> step chooseleaf firstn 0 type rack
> step
/dev/sdd
>>> [osd.14]
>>> host = store4
>>> osd journal = /dev/sdi1
>>> btrfs devs = /dev/sde
>>> [osd.15]
>>> host = store4
>>> osd journal = /dev/sdj1
>>> btrfs devs = /dev/sdf
>&
On Mon, Apr 1, 2013 at 2:16 PM, Sam Lang wrote:
> On Mon, Apr 1, 2013 at 5:59 AM, Papaspyrou, Alexander
> wrote:
>> Folks,
>>
>> we are trying to setup a ceph cluster with about 40 or so OSDs on our
>> hosting provider's infrastructure. Our rollout works with Opscode Chef, and
>> I'm driving my p
On Sat, Mar 30, 2013 at 3:46 AM, Wido den Hollander wrote:
> On 03/29/2013 01:42 AM, Steve Carter wrote:
>>
>> I create an empty 150G volume them copy it to a second pool:
>>
>> # rbd -p pool0 create --size 153750 steve150
>>
>> # /usr/bin/time rbd cp pool0/steve150 pool1/steve150
>> Image copy: 1
On Wed, Apr 3, 2013 at 9:45 AM, John Nielsen wrote:
> On Apr 1, 2013, at 3:33 PM, Gregory Farnum wrote:
>
>>> On Mon, Apr 1, 2013 at 2:16 PM, Sam Lang wrote:
>>>> On Mon, Apr 1, 2013 at 5:59 AM, Papaspyrou, Alexander
>>>> wrote:
>>>> 1. S
And if you put a big file in CephFS and then deleted it, the data will
be deleted from the RADOS cluster asynchronously in the background (by
the MDS), so it can take a while to actually get removed. :) If this
wasn't the behavior then a file delete would require you to wait for
each of those (10GB
On Fri, Apr 5, 2013 at 10:28 AM, Dimitri Maziuk wrote:
> On 04/05/2013 10:12 AM, Wido den Hollander wrote:
>
>> Think about it this way. You have two racks and the network connection
>> between them fails. If both racks keep operating because they can still
>> reach that single monitor in their ra
More details, please. You ran the same test twice and performance went
up from 17.5MB/s to 394MB/s? How many drives in each node, and of what
kind?
-Greg
Software Engineer #42 @ http://inktank.com | http://ceph.com
On Mon, Apr 8, 2013 at 12:38 PM, Ziemowit Pierzycki
wrote:
> Hi,
>
> I have a 3 n
If you were to do that you'd be going backwards in time with your
metadata, so — not really. CephFS is not generally production-ready at
this time, but we welcome bug reports!
-Greg
Software Engineer #42 @ http://inktank.com | http://ceph.com
On Mon, Apr 8, 2013 at 12:52 PM, Maik Kulbe
wrote:
>
On Wednesday, April 10, 2013 at 2:53 AM, Waed Bataineh wrote:
> Hello,
>
> I have several question i'll be appreciated if i got answers for them:
>
> 1. does the osd have a fixed size or it compatible with the machine
> i'm working with.
You can weight OSDs to account for different capacities or
Sounds like they aren't handling the transition very well when trying to
calculate old OSDs which might have held the PG. Are you trying to salvage the
data that was in it, or can you throw it away?
Can you post the backtrace they're producing?
-Greg
Software Engineer #42 @ http://inktank.com |
size 2
>
> that all.
>
>
>
>
>
>
> On 04/10/2013 05:24 PM, Gregory Farnum wrote:
> > Sounds like they aren't handling the transition very well when trying to
> > calculate old OSDs which might have held the PG. Are you trying to salvage
talij Poljatchek wrote:
> the log files.
>
> thank you ! :)
>
> On 04/10/2013 06:06 PM, Gregory Farnum wrote:
> > [Re-adding the list.]
> >
> > When the OSDs crash they will print out to their log a short description of
> > what happened, with a bunch of
Well, if you've made changes to your data which impacted the metadata,
and then you restore to a backup of the metadata pool, but not the
data, then what's there isn't what CephFS thinks is there. Which would
be confusing for all the same reasons that it is in a local
filesystem. You could construc
We've discussed the order of work (you can see my recent Ceph blog
post on the subject; though it's subject to revision) but haven't
committed to any dates at this time. Sorry. :(
-Greg
Software Engineer #42 @ http://inktank.com | http://ceph.com
On Wed, Apr 10, 2013 at 12:43 PM, Maik Kulbe
wrot
It's more or less a Ceph bug; the patch fixing this is in the 3.9-rc's
(although it should backport trivially if you're willing to build a
kernel: 92a49fb0f79f3300e6e50ddf56238e70678e4202). You can look at
http://tracker.ceph.com/issues/3793 if you want details.
-Greg
Software Engineer #42 @ http:/
That's certainly not great. Have you lost any data or removed anything
from the cluster? It looks like perhaps your MDS log lost an object,
and maybe got one shortened as well.
-Greg
Software Engineer #42 @ http://inktank.com | http://ceph.com
On Mon, Apr 8, 2013 at 11:55 PM, x yasha wrote:
> I'
Sam was complaining about this a few days ago too; I'm not sure if he
ever found a solution. If you've got the patience you can actually
build that image yourself; the source is in our git repo. If you
don't, I'm sure our intrepid doc/website people can get it fixed up
fairly soon. :)
-Greg
Softwar
"Incomplete" means that there are fewer than the minimum copies of the
placement group (by default, half of the requested size, rounded up).
In general rebooting one node shouldn't do that unless you've changed
your minimum size on the pool, and it does mean that data in those PGs
is unaccessible.
On Mon, Apr 15, 2013 at 2:42 AM, Olivier Bonvalet wrote:
> Hi,
>
> I have an OSD process which is regulary shutdown by scrub, if I well
> understand that trace :
>
> 0> 2013-04-15 09:29:53.708141 7f5a8e3cc700 -1 *** Caught signal
> (Aborted) **
> in thread 7f5a8e3cc700
>
> ceph version 0.5
On Mon, Apr 15, 2013 at 10:19 AM, Olivier Bonvalet wrote:
> Le lundi 15 avril 2013 à 10:16 -0700, Gregory Farnum a écrit :
>> Are you saying you saw this problem more than once, and so you
>> completely wiped the OSD in question, then brought it back into the
>> cluster, and
On Tue, Apr 16, 2013 at 8:47 AM, Elso Andras wrote:
> Hi,
>
> We are serving mp4 videos from cephfs, with lots of random read.
>
> Here is a graph: http://elbandi.net/ceph/ceph_bandwidth.png
> The issue is: there are +50% incoming bandwidth (osd -> cephfs mount, blue
> line) compared outgoing band
I believe Alex just merged format 2 reading into our testing branch, and is
working on writes now.
-Greg
On Thursday, April 18, 2013, Whelan, Ryan wrote:
> Does this mean its in linux-next? (released in 3.10?)
>
> - Original Message -
> From: "Olivier B." >
> To: "Ryan Whelan" >
> Cc: cep
What version was this on?
-Greg
On Thursday, April 18, 2013, Dan van der Ster wrote:
> Sorry for the noise.. we now have a better idea what happened here.
>
> For those that might care, basically we had one client looping while
> trying to list the / bucket with an incorrect key. rgw was handling
The barclamps were written against the crowbar "Betty" release, OpenStack
Essex (which is the last one supported by Crowbar), and Ceph "argonaut". JJ
has updated them to use "Bobtail", but I don't think anybody's run them
against newer versions of Openstack. :(
You should be able to find built vers
the volumes pool it fails.
> this is due to some permission failures probably because i try to follow a
> manual that is created for bobtail
> http://ceph.com/docs/master/rbd/rbd-openstack/
>
>
>
>
> On Thu, 2013-04-18 at 08:48 -0700, Gregory Farnum wrote:
>
> The
On Wed, Apr 17, 2013 at 7:40 AM, Guido Winkelmann
wrote:
> Hi,
>
> I just tried upgrading parts of our experimental ceph cluster from 0.56.1 to
> 0.60, and it looks like the new mon-daemon from 0.60 cannot talk to those from
> 0.56.1 at all.
>
> Long story short, we had to move some hardware aroun
Seeing this go by again it's simple enough to provide a quick
answer/hint — by setting the tunables it's of course getting a better
distribution of data, but the reason they're optional to begin with is
that older clients won't support them. In this case, the kernel client
being run; so it returns
ell.
>
>
> On Thu, Apr 18, 2013 at 1:56 PM, Gregory Farnum wrote:
>
>> Seeing this go by again it's simple enough to provide a quick
>> answer/hint — by setting the tunables it's of course getting a better
>> distribution of data, but the reason they're
Hey guys,
I finally had enough time to coordinate with a few other people and
figure out what's going on with the ceph-create-keys access denied
messages and create a ticket: http://tracker.ceph.com/issues/4752.
(I believe your monitor crash is something else, Matthew; if that
hasn't been dealt wit
On Thu, Apr 18, 2013 at 2:46 PM, Joao Eduardo Luis
wrote:
> On 04/18/2013 10:36 PM, Gregory Farnum wrote:
>>
>> (I believe your monitor crash is something else, Matthew; if that
>> hasn't been dealt with yet. Unfortunately all that log has is
>> messages, so it pr
There's a little bit of python called ceph-create-keys, which is
invoked by the upstart scripts. You can kill the running processes,
and edit them out of the scripts, without direct harm. (Their purpose
is to create some standard keys which the newer deployment tools rely
on to do things like creat
On Thu, Apr 18, 2013 at 7:59 PM, Mike Dawson wrote:
> Greg,
>
> Looks like Sage has a fix for this problem. In case it matters, I have seen
> a few cases that conflict with your notes in this thread and the bug report.
>
> I have seen the bug exclusively on new Ceph installs (without upgrading fro
On Fri, Apr 19, 2013 at 9:11 AM, Mike Dawson
wrote:
>
>
> On 4/19/2013 11:43 AM, Gregory Farnum wrote:
>>
>> On Thu, Apr 18, 2013 at 7:59 PM, Mike Dawson
>> wrote:
>>>
>>> Greg,
>>>
>>> Looks like Sage has a fix for this problem. In
On Fri, Apr 19, 2013 at 3:12 PM, Steven Presser wrote:
> Hey all,
> I've got a ceph cluster set up (0.56.4) on a custom centos image (base
> centos 6, plus kernel 3.6.9) running as a Xen dom0. I'm seeing a lot of
> messages like the ones at the bottom of this message. I'm entirely willing
>
tty early in the setup. (Heck, i just got it switched to
> gigabit ethernet...)
>
> Thanks,
> Steve
>
>
> On 04/19/2013 06:31 PM, Gregory Farnum wrote:
>>
>> On Fri, Apr 19, 2013 at 3:12 PM, Steven Presser
>> wrote:
>>>
>>> Hey all,
>>>
[ This is a good query for ceph-users. ]
On Sat, Apr 20, 2013 at 10:15 PM, James Harper
wrote:
> I'm doing some testing with ceph trying to figure out why my performance is
> so bad, and have noticed that there doesn't seem to be a way to cleanly stop
> an osd, or at least under debian /etc/ini
It's not entirely clear from your description and the output you've
given us, but it looks like maybe you've managed to bring up all your
OSDs correctly at this point? Or are they just not reporting down
because you set the "no down" flag...
In any case, CephFS isn't going to come up while the und
1301 - 1400 of 2358 matches
Mail list logo
