Am 21.06.2012 00:56, schrieb Sage Weil:
Just a quick update: there were some problems with doing a rolling
upgrade that may be responsible for these. We're testing the fix now.
Did this, by chance, happen on a cluster with a mix of 0.47.2 and 0.48?
No it was a clean install of the next branch
This is a bugfix release with one major fix and a few small ones:
* osd: disable use of the FIEMAP ioctl by default as its use was leading
to corruption for RBD users
* a few minor compile/build/specfile fixes
I was going to wait for v0.48, but that is still several days away. If
you are
One of the issues we have now with rgw is that it requires running a
maintenance utility every day so that we remove old objects. These
objects were left behind for a while, so that any pending read could
complete. There is no way currently to know whether there are reads in
progress on any object,
On 06/19/2012 11:46 PM, Stefan Priebe - Profihost AG wrote:
Am 19.06.2012 18:29, schrieb Dan Mick:
The number doesn't change currently (and can't currently be set
manually).
OK thanks, so for extending the storage this will be pretty important.
Do you have any plans for which version this fe
Just a quick update: there were some problems with doing a rolling
upgrade that may be responsible for these. We're testing the fix now.
Did this, by chance, happen on a cluster with a mix of 0.47.2 and 0.48?
sage
On Wed, 20 Jun 2012, Stefan Priebe wrote:
> Nobody an idea? Should i open up b
On 06/20/2012 09:09 AM, Sage Weil wrote:
On Wed, 20 Jun 2012, Stefan Priebe - Profihost AG wrote:
> Hello list,
>
> i've looked at the wiki (http://ceph.com/wiki/Ceph.conf) but there is no
exact
> desciption what filestore btrfs trans does. It seems it is false by default.
>
> Should it b
On Wed, 20 Jun 2012, Mandell Degerness wrote:
> The prior thread seems to refer to something fixed in 3.0.X, we are
> running 3.2.18. Also, in answer to the previous question, we see the
> error on systems running at 82% full and systems running at 5% full
> disks.
>
> Anyone have any ideas about
The prior thread seems to refer to something fixed in 3.0.X, we are
running 3.2.18. Also, in answer to the previous question, we see the
error on systems running at 82% full and systems running at 5% full
disks.
Anyone have any ideas about how to resolve the deadlock? Do we have
to configure mys
On Wed, Jun 20, 2012 at 2:53 PM, Travis Rhoden wrote:
> This incorrect syntax is still published in the docs at
> http://ceph.com/docs/master/ops/manage/crush/#adjusting-crush
>
> Are the docs that end up on that page in GitHub? I'd be happy to
> start making fixes and issuing pull-request. I've
This incorrect syntax is still published in the docs at
http://ceph.com/docs/master/ops/manage/crush/#adjusting-crush
Are the docs that end up on that page in GitHub? I'd be happy to
start making fixes and issuing pull-request. I've run into a few
gotchas when following the commands on those pag
I hit this a couple times and wondered the same thing. Why does the
OSD need to bail when it runs out of journal space?
On Wed, Jun 20, 2012 at 3:56 PM, Travis Rhoden wrote:
> Not sure if this is a bug or not. It was definitely user error -- but
> since the OSD process bailed, figured I would re
So, the docs for Chef install got some doc lovin' lately. It's all at
http://ceph.com/docs/master/install/chef/
http://ceph.com/docs/master/config-cluster/chef/
but the docs still stop short from having an actual running Ceph
cluster. Also, while the writing was in progress, I managed to lift
the
Am 20.06.2012 20:11, schrieb Sage Weil:
On Wed, 20 Jun 2012, Stefan Priebe wrote:
Am 20.06.2012 19:35, schrieb Sage Weil:
On Wed, 20 Jun 2012, Stefan Priebe wrote:
Nobody an idea? Should i open up bugs in tracker?
Let's open up bugs. If they are reproducible, debug osd = 20 logs would
be aw
Not sure if this is a bug or not. It was definitely user error -- but
since the OSD process bailed, figured I would report it.
I had /tmpfs mounted with 2.5GB of space:
tmpfs on /tmpfs type tmpfs (rw,size=2560m)
Then I decided to increase my journal size to 5G, but forgot to
increase the limit
On Wed, 20 Jun 2012, Travis Rhoden wrote:
> Is there a way I can monitor the utilization of the OSD journal? I
> would like to know if data is coming in fast enough to fill the
> journal, and thus slow down all the subsequent writes.
>
> Once the journal is full, that becomes the bottleneck for w
Is there a way I can monitor the utilization of the OSD journal? I
would like to know if data is coming in fast enough to fill the
journal, and thus slow down all the subsequent writes.
Once the journal is full, that becomes the bottleneck for writes, no?
(assuming that writes to the journal are
On Wed, 20 Jun 2012, Stefan Priebe wrote:
> Am 20.06.2012 19:35, schrieb Sage Weil:
> > On Wed, 20 Jun 2012, Stefan Priebe wrote:
> > > Nobody an idea? Should i open up bugs in tracker?
> >
> > Let's open up bugs. If they are reproducible, debug osd = 20 logs would
> > be awesome!
> >
> > Also,
Am 20.06.2012 19:35, schrieb Sage Weil:
On Wed, 20 Jun 2012, Stefan Priebe wrote:
Nobody an idea? Should i open up bugs in tracker?
Let's open up bugs. If they are reproducible, debug osd = 20 logs would
be awesome!
Also, the crash you mentioned in your earlier email we did see:
htt
On Wed, 20 Jun 2012, Stefan Priebe wrote:
> Nobody an idea? Should i open up bugs in tracker?
Let's open up bugs. If they are reproducible, debug osd = 20 logs would
be awesome!
Also, the crash you mentioned in your earlier email we did see:
http://tracker.newdream.net/issues/2599
If
Nobody an idea? Should i open up bugs in tracker?
Am 20.06.2012 15:30, schrieb Stefan Priebe - Profihost AG:
Mhm always the same osd's are crashing now again. Mostly while shutting
down or restarting a KVM machine.
This time:
### Server 1
0> 2012-06-20 11:59:0
On Wed, Jun 20, 2012 at 9:46 AM, Stefan Priebe wrote:
>> Actually, the short hostname of the machine is completely unrelated to
>> what IP addresses etc it has. The hostname is a configurable string.
>> We treat it as such -- it's just expected to be a unique identifier
>> for the host.
> Correct.
Am 20.06.2012 um 18:39 schrieb Tommi Virtanen :
> On Wed, Jun 20, 2012 at 2:20 AM, Stefan Priebe - Profihost AG
> wrote:
>>> What does hostname | cut -d . -f 1 say?
>>
>> It says ssdstor000i as the machine defaults to it's public IP.
>>
>> Whouldn't it make sense to go through all aliases set
On Wed, Jun 20, 2012 at 2:20 AM, Stefan Priebe - Profihost AG
wrote:
>> What does hostname | cut -d . -f 1 say?
>
> It says ssdstor000i as the machine defaults to it's public IP.
>
> Whouldn't it make sense to go through all aliases set in /etc/hosts which
> point to a local IP?
Actually, the sho
On Wed, 20 Jun 2012, Stefan Priebe - Profihost AG wrote:
> Hello list,
>
> i've looked at the wiki (http://ceph.com/wiki/Ceph.conf) but there is no exact
> desciption what filestore btrfs trans does. It seems it is false by default.
>
> Should it be enabled when using btrfs?
Nope. This is the o
On Wed, 20 Jun 2012, Guan Jun He wrote:
>
>
> >>> On 6/19/2012 at 11:33 PM, in message
> , Sage Weil
> wrote:
> > [Sorry for not responding earlier!]
> >
> > On Tue, 19 Jun 2012, Guan Jun He wrote:
> >> Hi,
> >>
> >> Do you think this is needed?
> >> The osdmap update need to hold this sem
On 06/20/2012 11:20 AM, Stefan Priebe - Profihost AG wrote:
Am 20.06.2012 11:13, schrieb Wido den Hollander:
On 06/20/2012 10:59 AM, Stefan Priebe - Profihost AG wrote:
Hello list,
i've some problems to find the correct combination of ceph.conf,
/etc/hosts for the init script to work correctly
Mhm always the same osd's are crashing now again. Mostly while shutting
down or restarting a KVM machine.
This time:
### Server 1
0> 2012-06-20 11:59:06.499813 7f1664052700 -1 *** Caught signal
(Segmentation fault) **
in thread 7f1664052700
ceph version 0.4
Hello list,
i'm still seeing osd crashes with next branch under KVM load. If you
need the core dump please tell me.
Here are TWO different crashes.
Here are the last log lines:
### CRASH 1 ###
-3> 2012-06-20 11:59:06.446836 7f1660f4b700 0 osd.13 105 pg[4.64b(
v 105'297
>>> On 6/19/2012 at 11:33 PM, in message
, Sage Weil
wrote:
> [Sorry for not responding earlier!]
>
> On Tue, 19 Jun 2012, Guan Jun He wrote:
>> Hi,
>>
>> Do you think this is needed?
>> The osdmap update need to hold this sem.As in function
>> void ceph_osdc_handle_map(struct ceph_osd_cl
Am 20.06.2012 11:20, schrieb Stefan Priebe - Profihost AG:
Am 20.06.2012 11:13, schrieb Wido den Hollander:
On 06/20/2012 10:59 AM, Stefan Priebe - Profihost AG wrote:
Hello list,
i've some problems to find the correct combination of ceph.conf,
/etc/hosts for the init script to work correctly.
Am 20.06.2012 11:13, schrieb Wido den Hollander:
On 06/20/2012 10:59 AM, Stefan Priebe - Profihost AG wrote:
Hello list,
i've some problems to find the correct combination of ceph.conf,
/etc/hosts for the init script to work correctly.
My systems are named:
ssdstor000i
ssdstor001i
ssdstor002i
On 06/20/2012 10:59 AM, Stefan Priebe - Profihost AG wrote:
Hello list,
i've some problems to find the correct combination of ceph.conf,
/etc/hosts for the init script to work correctly.
My systems are named:
ssdstor000i
ssdstor001i
ssdstor002i
What does hostname | cut -d . -f 1 say?
This i
Hello list,
i've some problems to find the correct combination of ceph.conf,
/etc/hosts for the init script to work correctly.
My systems are named:
ssdstor000i
ssdstor001i
ssdstor002i
These names point to the public external IP.
These systems also have an alias (ssdstor000,ssdstor001) point
33 matches
Mail list logo
