For all of these use cases you're going to want to flush the journal
to permanent store to make sure you don't lose any data.
1) Shut down the cosd.
2) Start it up again with the --flush-journal argument.
It will run in the foreground and flush out any data in the journal.
You're all set and can
What version of the code are you using? Master was broken for a couple
hours today but should be fixed now.
On Thu, Jun 9, 2011 at 6:17 PM, Dyweni - Ceph-Devel
ys3fpfe2y...@dyweni.com wrote:
Hi All,
Just tripped over this bug:
*** Caught signal (Segmentation fault) **
in thread 0xb6fff6d0
2011/6/6 Székelyi Szabolcs szeke...@niif.hu:
Hi all,
I have a three node ceph setup, two nodes playing all three roles (OSD, MDS,
MON), and one being just a monitor (which happens to be the client I'm using
the filesystem from).
I want to achieve high availablity by mirroring all data
2011/6/2 biyan chen riby.c...@gmail.com:
2011-06-02 16:27:35.010776 7f3f53940720 unable to authenticate as client.admin
That's your problem. You don't have your keys set up properly to
authenticate as the admin on those nodes. :)
I suspect permissions problems are the cause of your other issues
/]# ll
total 112
drwxr-xr-x. 1 root root 0 Jun 1 20:05 1
I to execute the command :
mount -o remount, rw /1
Still hint without permission
2011/5/27 Gregory Farnum gr...@hq.newdream.net:
I'm not quite sure what you're saying here -- can you give a more
precise description
Mounting a Ceph cluster isn't quite like mounting an NFS machine. You
mount by connecting to one of the monitor servers, and then you
dynamically connect and disconnect to the other machines as needed. :)
So if you can only connect to your first server, using the monitor
address, then that's
This is an interesting one -- the invariant that assert is checking
isn't too complicated (that the object lives on the RecoveryWQ's
queue) and seems to hold everywhere the RecoveryWQ is called. And the
functions modifying the queue are always called under the workqueue
lock, and do maintenance if
You are really putting this stuff through the wringer -- nobody else
hits these issues so frequently!
Can you remind us of your test setup? I'd like to figure out how
you're killing these FSes so badly.
And just randomly, I'd try ext4 before xfs. :)
-Greg
On Fri, May 27, 2011 at 4:59 AM, Fyodor
On Fri, May 27, 2011 at 9:06 AM, Sylar Shen kimul...@gmail.com wrote:
Hi developers,
Recently I am testing radosgw about its functions.
Then I am curious that will radosgw implement the function or API of
quota in the future?
I've checked the mail archives and found that the quota issue had
On Thu, May 19, 2011 at 1:27 AM, Dyweni - Ceph-Devel
ys3fpfe2y...@dyweni.com wrote:
4 x OSD
OSD1 =
P3 1Ghz, 256K Cache
2GB RAM
80GB IDE (dd_rescue can read about 45MB/s sustained)
1Gb NIC (e1000 driver)
OSD2 =
P3 550Mhz, 512K Cache
768MB RAM
40GB IDE (dd_rescue can
On Thu, May 19, 2011 at 3:56 AM, Qian Wei hzqian...@gmail.com wrote:
Hello friends!
I am learning ceph. Does anyone use IDE tool for Ceph development?
Can you tell me something about IDE for ceph? Thanks very much for
your reading my e-mail.
I use Eclipse in my day-to-day work on Ceph, and it
On Fri, May 20, 2011 at 3:05 AM, Henry C Chang henry.cy.ch...@gmail.com wrote:
Hi all,
In my ceph cluster, I saw a pg membership change as below:
At osdmap epch 160, pg 1.184 was mapped to [4, 6].
osdmaptool: osdmap file '160'
parsed '1.184' - 1.184
1.184 raw [4,6]
On Fri, May 20, 2011 at 10:59 AM, Fyodor Ustinov u...@ufm.su wrote:
Hi!
How should I respond to the message libceph: get_reply unknown tid 68724
from osd0 in syslog?
That means something in the messaging layers got a bit busted, and you
should generally respond by filing bug reports. :) What
I'm not quite sure what you're saying here -- can you give a more
precise description of the problem?
On Tue, May 24, 2011 at 12:44 AM, biyan chen riby.c...@gmail.com wrote:
mkcephfs -c /usr/local/etc/ceph/ceph.conf --allhosts --mkbtrfs -k
/usr/local/etc/ceph/keyring.bin
/etc/init.d/ceph -a
On Thu, May 26, 2011 at 3:30 PM, Fyodor Ustinov u...@ufm.su wrote:
On 05/26/2011 08:37 PM, Gregory Farnum wrote:
That means something in the messaging layers got a bit busted, and you
should generally respond by filing bug reports.:) What version are
you on right now? I think we fixed some
On Sun, May 15, 2011 at 11:44 PM, Simon Tian aixt2...@gmail.com wrote:
Great! Thx very much...
It's just that reason!!!
I just did rados -p data bench 60 write -t 1 -b 1024, the speed is
more or less the same.
So I changed the IO_SIZE to 1MB and above in my test codes, the speed
became
Did you actually re-build the code? Configure just sets up some of the scripts:
1) ./configure #(with desired environment settings)
2) ./autogen.sh
3) make #(hopefully -j8 or something)
Then you'll need to make sure that the built library is the one you're
linking with -- how to do this varies a
On Fri, May 13, 2011 at 1:00 AM, Sylar Shen kimul...@gmail.com wrote:
Hi developers,
I have a question between the numbers of cosd and memory cache in a server.
Here is my environment
I got 15 servers as OSDes and each one has 8 cores CPU, 16GB RAM and
1TB*12 Disks.
I only have 1 MDS and
So it occurs to me that one thing we could do in the Objecter layer,
to make this a bit more sane, is to optionally enable caching of the
missing objects. Then the OSD could send out all the missing objects
with a flag saying that they're missing, the Objecter could cache this
list, go through the
be
misinterpreting that. I can look at that again.
On Tue, May 10, 2011 at 12:44 PM, Gregory Farnum gr...@hq.newdream.net
wrote:
What are the 'l'-prefixed functions for? They look like they're just
duplicates except for the prefix.
-Greg
Ah, right. I didn't notice the different flag
On Tue, May 10, 2011 at 1:34 PM, Yehuda Sadeh Weinraub
yehud...@gmail.com wrote:
lsetxattr and friends are known linux API functions, and have
corresponding system calls with the same name. I say let's keep them
as is.
Heh. Should probably just ignore clueless me in that case. :)
--
To
Ermm, doesn't rados sync do a whole tree sync or something? I think we
want to maintain a simple get/set object capability and the rados tool
is a pretty logical place for that.
-Greg
On Mon, May 9, 2011 at 3:21 PM, Colin McCabe
cmcc...@ceph-tracker.dreamhostps.com wrote:
The branch, master has
Ermm, doesn't rados sync do a whole tree sync or something? I think we
want to maintain a simple get/set object capability and the rados tool
is a pretty logical place for that.
-Greg
On Mon, May 9, 2011 at 3:21 PM, Colin McCabe
cmcc...@ceph-tracker.dreamhostps.com wrote:
The branch, master has
On Mon, May 9, 2011 at 4:24 PM, Gregory Farnum
gregory.far...@dreamhost.com wrote:
Ermm, doesn't rados sync do a whole tree sync or something? I think we
want to maintain a simple get/set object capability and the rados tool
is a pretty logical place for that.
-Greg
Sorry, Gmail got confused
Ah, yeah, it sounds like you broke your mon map by trying to change
the name of your active monitor. I'm pretty sure to make that work you
would need to add the monitor under the new name and then remove the
old name!
Let us know if you run into any other trouble, you're probably
touching a lot
On Sun, May 8, 2011 at 7:43 PM, Simon Tian aixt2...@gmail.com wrote:
Hi folks,
I am testing the replication performance of ceph-0.26 with
libceph, write 1G data in with ceph_write() and read it out with
ceph_read(),
rep_size 1 2
3
On Sun, May 8, 2011 at 8:04 PM, Simon Tian aixt2...@gmail.com wrote:
For primary copy, I think when the replication size is 3, 4, or even
more, the writing speed should also near with 2 replication. Because
the 2nd, 3rd, 4th, ... replication are written parallelly. The speed
I got for 3, 4
Nope, the MDS is only required for the POSIX-compliant Ceph layer.
-Greg
On Thursday, May 5, 2011 at 10:16 AM, Fyodor Ustinov wrote:
Hi!
If I only want to use rbd - requires metaserver?
WBR,
Fyodor.
--
To unsubscribe from this list: send the line unsubscribe ceph-devel in
the body
On Thu, Apr 28, 2011 at 4:55 PM, Zenon Panoussis ora...@provocation.net wrote:
On 04/28/2011 10:02 PM, Gregory Farnum wrote:
[various explanations]
Thanks Greg, that's very helpful towards graspings ceph's workings. I'll
put it in the wiki.
The relation between these reports and your data
On Thu, Apr 28, 2011 at 4:55 PM, Zenon Panoussis ora...@provocation.net wrote:
On 04/28/2011 10:02 PM, Gregory Farnum wrote:
1) You have 9791 MB of data in the filesystem.
2) You have (1MB - 9791 MB=) 2431MB of metadata maintaining the Ceph
tree.
3) RADOS is using 2MB of disk space
On Thursday, April 28, 2011 at 11:51 AM, Zenon Panoussis wrote:
What is the current default object size? A paper from 2004 by Sage
et al speaks of 1 MB, a later one of 8 MB and
http://diaspora.gen.nz/~rodgerd/archives/1219-Ceph.html says 4 MB.
Is there a way to configure it? And is there any
On Wednesday, April 27, 2011 at 12:42 PM, Fabio Kaminski wrote:
Ok, thats the way it should be.. :)
But specializing a little more the question.. whats the data
partition scheme between nodes, how can a user control it.. block
level? file level?
suppose that i have agregations that i want it
Looking at what we have here, we are unlikely to need:
auth, paxos, monc
above 5 (or probably 1). The code these run is pretty stable and hasn't
exhibited significant issues in a long time IIRC.
Unfortunately the MDS is one we're likely to want to keep high, but I suppose
that with debug mds 10
I really don't see how pushing the naming complexity into the local filesystem,
where it adds lots of otherwise-useless inodes and dentries, is going to help
us.
I like what Yehuda has here for its relative simplicity -- though I think we
should just up the hash size enough that we don't need
On Wednesday, April 20, 2011 at 12:02 PM, Colin McCabe wrote:
Hi guys,
I pushed a branch with a revised libceph API. The main changes:
* similar appearance to librados. It should be intuitive if you've
used librados.
* The new API allows multiple ceph clusters to be open at once
(although
On Tue, Apr 19, 2011 at 3:45 AM, Zenon Panoussis ora...@provocation.net wrote:
It's a bit more complicated than that. While we could probably do a better
job of controlling bandwidths, there are a lot of pieces devoted to handling
changes in disk performance and preventing the OSD from
I looked through your logs a bit and noticed that the OSD on node01 is crashing
due to high latencies on disk access (I think the defaults for this case are it
asserts out if there's no progress after 10 minutes or something).
Based on that, I pretty much have to guess that there's just too
On Monday, April 18, 2011 at 3:38 PM, Zenon Panoussis wrote:
On 04/18/2011 11:21 PM, Gregory Farnum wrote:
I looked through your logs a bit and noticed that the OSD on node01 is
crashing due to high latencies on disk access (I think the defaults for
this case are it asserts out
On Thu, Apr 14, 2011 at 8:47 PM, Qian Wei hzqian...@gmail.com wrote:
Hello everyonel:
Recently, I have put up a simple test platform. Now I want to test
the MDS cluster's performance,but I don't how to get it. Can anyone
give me some advices or some tools may help me get it.
Any kind
I looked at your logs and they're very interesting. I suspect the reason your
journal broke is because your logging partition filled up, although I can't be
sure.
We have seen a similar error once before but were unable to diagnose it because
there wasn't enough logging, and it looks like we
On Thu, Apr 14, 2011 at 2:46 AM, huang jun hjwsm1...@gmail.com wrote:
hi developers,
I have a question about the prealloc_inos variable in SessionMap.h
when the client first request to allocate an inode,
prepare_new_inode function will prepare new inodes to the session. so
if the
On Wednesday, April 13, 2011 at 5:12 AM, Martin Wilderoth wrote:
The statistics of the disks, this is after the crash of osd2 and osd4.
/dev/sdc 143373312 124954676 18418636 88% /data/osd0
/dev/sdd 143373312 137639524 5733788 97% /data/osd1
/dev/sdc 143373312 120350584 23022728 84%
On Tuesday, April 12, 2011 at 11:05 AM, Martin Wilderoth wrote:
Thanks for the answer, now I know the reson. Some of my osd had 90% of data,
dmesg also shows error with the btrfs on the hosts. I will run the test with
another file system ext3 :-) or is any other filesystem better. It's a
On Fri, Apr 8, 2011 at 4:28 PM, Fyodor Ustinov u...@ufm.su wrote:
Hi.
Looks as if mds uses only 1 cpu on the connection. This is so, or I made a
mistake somewhere in the settings?
This is basically right. The MDS has a bit of threading but most
operations depend on a global lock so it's
On Monday, April 11, 2011 at 8:35 AM, AnnyRen wrote:
Hi,
I want to test the reliability of MDS in Ceph File System,
there are three MONs, two MDSes and three OSD in my test environment.
With ceph -s, I saw the MDS's status is
2011-04-11 23:30:38.001177 mds e5: 1/1/1 up {0=up:active}, 1
Unfortunately a lot of the config options aren't really defined or
described anywhere outside the code. Some of them are described in the
wiki at http://ceph.newdream.net/wiki/Ceph.conf, and possibly on other
pages (use the search option).
Feel free to ask about specific options you're interested
On Thursday, April 7, 2011 at 5:21 PM, Colin McCabe wrote:
I'm also aware that injectargs changes configuration values that other
threads may be reading, without using a lock or an atomic update. So
far, this doesn't seem to have created any problems, but it is
technically incorrect and a
On Mon, Apr 4, 2011 at 11:44 PM, Rutger ter Borg rut...@terborg.net wrote:
thanks for the elaborate answer! With respect to the compile error, I got
this in an empty file with just including rados/librados.hpp
$ g++ ./rados.cpp
In file included from /usr/include/rados/buffer.h:55:0,
On Mon, Apr 4, 2011 at 4:33 AM, Rutger ter Borg rut...@terborg.net wrote:
Hello,
I'm in the progress of evaluating librados as an object store. I'm using
Debian's latest packages as of today, and noted that I need to define
NO_ATOMIC_OPS to get something compiled that includes librados.hpp.
On Monday, April 4, 2011 at 2:55 PM, Fyodor Ustinov wrote:
Hi.
ceph osd pool set data size 1
dd if=/dev/zero of=aaa bs=1024000 count=4000
409600 bytes (4.1 GB) copied, 31.3153 s, 131 MB/s
ceph osd pool set data size 2
409600 bytes (4.1 GB) copied, 72.7146 s, 56.3 MB/s
ceph
On Monday, April 4, 2011 at 4:28 PM, Fyodor Ustinov wrote:
On 04/05/2011 01:52 AM, Gregory Farnum wrote:
Data replicas don't service reads unless you specifically configure it for
that (it doesn't provide POSIX semantics if you do that), and each replica
needs to get the data written
On Sat, Apr 2, 2011 at 3:55 AM, Martin Wilderoth
martin.wilder...@linserv.se wrote:
Hello,
I have seperate partitions for my osd and the btrfs file system.
I also use SSD-disk for journaling.
But I got problem when the root system was filled up with logfiles on one
host,
the file system
I think that's the user_xattr error (if you don't have user xattrs enabled on
the mount).
If not there were some issues on master yesterday or two days ago that I think
are fixed now; somebody else should know more about what those were.
-Greg
On Thursday, March 31, 2011 at 10:59 AM, Noah
are archived and indexed by search engines so other people can
get their answer by google instead of email! :)
Thanks .
- Original Message -
From: Gregory Farnum gregory.far...@dreamhost.com
To: huang jun hjwsm1...@gmail.com
Cc: ceph-devel@vger.kernel.org
Sent: Thursday, March 24, 2011 10:27
Hi Wilfrid,
If you used a promo code you should be able to create a Ceph Object Storage
user. This will give you access to the playground via our S3 lookalike service,
which you can set up to serve images the same way you set up Amazon S3 to do so.
You can also create users to use the Ceph
2011/3/23 huang jun hjwsm1...@gmail.com:
hi,
yes,the mds is overloaded. there are four clients in the ceph
cluster,and each of them send a great many of requests to the mds. i
think the standby should become active and share the workload rather
than take over the laggy one in this case. what
On Sat, Mar 19, 2011 at 11:43 PM, Martin Wilderoth
martin.wilder...@linserv.se wrote:
I have a small ceph cluster with 4 osd ( 2 disks on 2 hosts).
I have been adding and removing files from the file system, mounted as ceph
on an other host.
Now I have removed most of the data on the file
You need to include the port to connect to. By default this will be 6789:
[root@ceph86 /]# mount -t ceph 10.65.1.86:6789:/ /mnt/ceph
That is a pretty terrible error message though; I'll put a note in the tracker!
It's also possible you have other issues based on that libceph error
(did you
On Mon, Mar 21, 2011 at 9:22 AM, Colin McCabe cmcc...@alumni.cmu.edu wrote:
The more annoying dependencies are things like GTK2, which don't
really affect the behavior of the rest of the project, but which will
have to be mandatory in the RPM, it seems. Also, as I said, tcmalloc
will present
On Monday, March 21, 2011 at 10:32 AM, Ben De Luca wrote:
Sorry to jump into the converstation, how slow can the deletion of
files actually be?
One of the tests I ran a few weeks ago had me generating files,
deleting them and then writing them again from a number of clients. I
noticed that
:-)
- Ursprungligt meddelande -
Från: Ben De Luca bdel...@gmail.com
Till: Gregory Farnum gregory.far...@dreamhost.com
Kopia: Martin Wilderoth martin.wilder...@linserv.se,
ceph-devel@vger.kernel.org
Skickat: måndag, 21 mar 2011 18:32:46
Ämne: Re: Disk allocation
On Monday, March 21, 2011 at 4:43 AM, Wido den Hollander wrote:
Hi,
On Sun, 2011-03-20 at 13:51 -0500, Mark Nigh wrote:
Wido,
Thanks for your input. Has there be any talk/discussions around support on
the upcoming release of v1.0. I see in the roadmap that is slated for
April. I
Yeah, we put in a ticket for this a couple days ago and it's being worked on. :)
-Greg
On Fri, Mar 18, 2011 at 3:24 AM, Lars Täuber taeu...@bbaw.de wrote:
Hello developpers,
please don't mind me writing to this list.
I can't find a contact for the one responsible for the wiki.
Can you tell
On Friday, March 18, 2011 at 3:45 PM, Tommi Virtanen wrote:
On Fri, Mar 18, 2011 at 10:23:37PM +, Steena Monteiro wrote:
I'm trying to build the Ceph client on Ubuntu 10.04.2 with kernel
2.6.32-28-generic. make -C install gives:
/usr/src/ceph-client-standalone/ceph/super.c:870:
On Friday, March 18, 2011 at 3:50 PM, Gregory Farnum wrote:
On Friday, March 18, 2011 at 3:45 PM, Tommi Virtanen wrote:
On Fri, Mar 18, 2011 at 10:23:37PM +, Steena Monteiro wrote:
I'm trying to build the Ceph client on Ubuntu 10.04.2 with kernel
2.6.32-28-generic. make -C
On Fri, Mar 11, 2011 at 6:31 AM, Ajit K Jena a...@iitb.ac.in wrote:
Hi Greg,
I found that the mds process dies during the weird 'untar'
behavior. I am trying to analyze the logs. May be I will
find something there.
Can you send us the backtrace? It should be printed out at the end of
the mds
On Friday, March 11, 2011 at 10:51 AM, Jim Schutt wrote:
I'm also confused. Here's the code I ran:
void put() {
//generic_dout(0) RefCountedObject::put this nref.read()
- (nref.read() - 1) dendl;
if (nref.dec() == 0) {
utime_t s = g_clock.now();
delete this;
utime_t e
Hmmm, your conf file isn't obviously broken but you shouldn't be having any
trouble with an untar, either.
When the untar hangs, have you checked to make sure none of the daemons have
crashed? Are they all idle, or is one of them using a lot of CPU?
-Greg
On Thursday, March 10, 2011 at 3:04 AM,
Hmm, this does seem odd. How do you have Ceph installed and configured?
Just to be clear:
You have one machine hosting a monitor, MDS, and OSD. You reboot that machine
and the MDS and monitor come back up, but the OSD doesn't. And you appear to
have an OSD running somewhere else?
This doesn't
2011/3/10 huang jun hjwsm1...@gmail.com:
hi Greg:
here is my configure file:
snip
and i find something wrong in /var/log/ceph/osd.0.log
2011-03-11 08:42:23.888496 7f207255d6f0 osd0 0 mounting /data/osd0
2011-03-11 08:42:23.888517 7f207255d6f0 filestore(/data/osd0) basedir
/data/osd0 journal
On Wednesday, March 9, 2011 at 8:02 AM, Jim Schutt wrote:
Heh. Evidently I didn't do the instrumentation right.
Or more specifically, I didn't look in the right places
for the result I needed.
Once I understood the code/logging enough to write a
script to look for delayed osd_ping message
On Wednesday, March 9, 2011 at 10:36 AM, Jim Schutt wrote:
Here's another example with more debugging. The
PG count during this interval is:
2011-03-09 10:35:58.306942 pg v379: 25344 pgs: 25344 active+clean; 12119 MB
data, 12025 MB used, 44579 GB / 44787 GB avail
2011-03-09
Sylar:
Did you run this using Ceph's FUSE or in-kernel client? If this is on cfuse,
the results don't surprise me -- it's not very well optimized!
If this is using the kernel client, I'd expect much faster results on the
buffered write test -- I think speeds approaching the network interface
On Tuesday, March 8, 2011 at 12:52 AM, Upendra Moturi wrote:
Hi
I am having issue witrh replication levels.
Even though i have set data @2x and mds@1x replication,the data copied
is occupying 4x space
How are you measuring the data use as being at 4x? Keep in mind that data will
be
On Friday, March 4, 2011 at 11:48 AM, Sage Weil wrote:
Hi,
I'm continuing my testing of the master branch
(commit 1ed2d8c587) against 96 osds.
I'm having trouble getting a new filesystem to
start up. FWIW this size filesystem starts up
in a minute or two at most under the stable
On Friday, March 4, 2011 at 12:08 PM, Colin McCabe wrote:
tl;dr: cpu_profiler may randomly crash your program on x86_64.
The problem should be resolved in the future.
From what I've read, it's unfortunately not out of date. However, it manifests
itself as a crash so it's not the reason that
On Friday, March 4, 2011 at 12:53 PM, Yehuda Sadeh Weinraub wrote:
you mean 'osd pg bits' in the ceph.conf?
Sorry, yes.
Jim, that 3 second gap is a configurable timeout/retry period; I don't remember
exactly where it's set up. I suspect this failed reads problem is some kind of
fallout from
On Wednesday, March 2, 2011 at 3:25 PM, Jim Schutt wrote:
On Wed, 2011-03-02 at 16:20 -0700, Gregory Farnum wrote:
On Wed, Mar 2, 2011 at 2:57 PM, Jim Schutt jasc...@sandia.gov wrote:
On Wed, 2011-03-02 at 14:59 -0700, Jim Schutt wrote:
On Wed, 2011-03-02 at 14:45 -0700, Sage
On Saturday, February 26, 2011 at 11:52 AM, Gregory Farnum wrote:
On Saturday, February 26, 2011 at 11:40 AM, Laszlo Boszormenyi wrote:
On Sat, 2011-02-26 at 10:28 -0800, Gregory Farnum wrote:
This looks like an error that came up (and got fixed) on Ubuntu several
months ago:
https
On Saturday, February 26, 2011 at 3:41 AM, Laszlo Boszormenyi wrote:
Hi List,
Being the Debian packager of Ceph, I ask for help about building 0.24.3
on armel. It fails with:
g++ -DHAVE_CONFIG_H -I. -Wall -D__CEPH__ -D_FILE_OFFSET_BITS=64
-D_REENTRANT -D_THREAD_SAFE -rdynamic -g -O2 -MT
On Saturday, February 26, 2011 at 11:40 AM, Laszlo Boszormenyi wrote:
On Sat, 2011-02-26 at 10:28 -0800, Gregory Farnum wrote:
This looks like an error that came up (and got fixed) on Ubuntu several
months ago:
https://bugs.launchpad.net/ubuntu/+source/ceph/+bug/635388
Yes, it looks
On Thursday, February 24, 2011 at 6:19 PM, Colin McCabe wrote:
The branch, librados_api has been updated
via d15fe7f90bf488c6251be4633b104a822ebd134f (commit)
from e3f3177febf07bf3b311414167fdf4ccbf07c472 (commit)
Those revisions listed above that are new to this repository have
not
Sorry, we kind of lost this in the shuffle!
Did you manage to resolve this issue? From what I see in your steps below, it
looks like you're doing things in the wrong order -- mkcephfs should always be
the first thing you do. If you really have nodes running before you do mkcephfs
I suspect you
On Wednesday, February 23, 2011 at 7:57 AM, H Chang wrote:
Hi,
I was browsing PG management code in Ceph, and have one question.
It appears that an OSD could send a PG removal request to another OSD.
The recipient OSD would then remove that particular PG from its pg_map.
I'd like to know
On Wednesday, February 23, 2011 at 9:52 AM, Jim Schutt wrote:
On Thu, 2011-02-17 at 09:11 -0700, Sage Weil wrote:
On Thu, 17 Feb 2011, Jim Schutt wrote:
Hi Sage,
On Wed, 2011-02-16 at 17:54 -0700, Sage Weil wrote:
On Wed, 16 Feb 2011, Sage Weil wrote:
shouldn't affect
On Wednesday, February 23, 2011 at 10:54 AM, Sage Weil wrote:
On Wed, 23 Feb 2011, Gregory Farnum wrote:
I have managed to get OSDs wrongly marking each other down during
startup when they're peering large numbers of PGs/pools, as they
disagree on who they need to be heartbeating (due
On Wednesday, February 23, 2011 at 11:23 AM, Jim Schutt wrote:
I have managed to get OSDs wrongly marking each other down during startup
when they're peering large numbers of PGs/pools, as they disagree on who
they need to be heartbeating (due to the slow handling of new osd maps and
pg
On Thursday, February 17, 2011 at 11:13 PM, Sage Weil wrote:
On Thu, 17 Feb 2011, Jim Schutt wrote:
Why should it take 28 seconds to add a new timer event?
Huh.. that is pretty weird. I see multiple sync in there, too, so it's
not like something was somehow blocking on a btrfs commit.
On Friday, February 18, 2011 at 5:30 AM, Kiran Patil wrote:
Hello,
May I know when would it be possible to use the underlying btrfs
features such as dedupe(offline) and compression for ceph cosds ?
Is it possible to access btrfs features in cosd seamlessly ?
Well, there are two ways to run
On Wednesday, February 16, 2011 at 1:25 PM, Jim Schutt wrote:
Hi,
I've been testing v0.24.3 w/ 64 clients against
1 mon, 1 mds, 96 osds. Under heavy write load I
see:
[WRN] map e7 wrongly marked me down or wrong addr
I was able to sort through the logs and discover that when
this
On Monday, February 14, 2011 at 5:37 AM, Longguang Yue wrote:
osd e35: 3 osds: 3 up, 3 in -- 2 blacklisted MDSes
ceph can recover blacklisted osds by itself, but I want to recover blacklist
osds as soon as possible.
How to use the command, 'ceph osd blacklist rm host:port'
On Mon, Feb 7, 2011 at 1:31 PM, Jim Schutt jasc...@sandia.gov wrote:
On Mon, 2011-02-07 at 14:04 -0700, Yehuda Sadeh Weinraub wrote:
It does look like this specific problem.
You can try cherry-pick commit 9c3db35 off the ceph git. It is just a
temporary workaround, and it wasn't tested too
On Tue, Feb 1, 2011 at 9:27 AM, Tommi Virtanen
tommi.virta...@dreamhost.com wrote:
That has been in the plans from day one. Line branch coverage are
awesome tools.
The first time I used Clover I think my team and I spent 2 hours
picking our jaws up off the floor. They don't really discuss that
On Mon, Jan 31, 2011 at 9:01 PM, DongJin Lee dongjin@auckland.ac.nz wrote:
I'm using the unstable version dated 20th-Jan.
When I was starting up multiple OSDs, e.g., 3 or more.
iostat shows, after the start, the OSDs utilizes to 100% themselves
(there's not much of traffic going on)
If
2011/1/31 Cláudio Martins c...@ist.utl.pt:
So, I do agree that, as Brian Chrisman pointed out, background scrub
is always important as it helps to prevent your data redundancy going
bad without you knowing about it. I also agree with that sys. admin.
notification is important in either case.
On Wed, Jan 26, 2011 at 5:36 PM, Longguang Yue
longguang_...@tcloudcomputing.com wrote:
I update 0.23.1 to 0.24.2. I do not success in running ceph 0.24*
version.!!
L meet lots of error. This is step:
1.make ceph.rpm ---ok
2.mkcephfs ;first time is
On Sun, Jan 23, 2011 at 6:07 AM, Chris Webb ch...@arachsys.com wrote:
One feature I would really like to be able to export to users is an ability
to make copy-on-write clones of virtual hard drives, in a Ceph context
generating a new rbd image from an existing one, or from a snapshot of an
On Sun, Jan 23, 2011 at 8:28 PM, Longguang Yue
longguang_...@tcloudcomputing.com wrote:
2 out of 3 is down,so the live osd is full, data lose.
How to solve the case?
Well, there's no way to magically make Ceph store more data than the
OSD can hold. However, if you can bring the other OSDs back
On Thu, Jan 20, 2011 at 9:13 AM, tsuna tsuna...@gmail.com wrote:
On Tue, Jan 18, 2011 at 11:54 AM, Gregory Farnum gr...@hq.newdream.net
wrote:
Are you running on 32-bit machines? It looks like we're accidentally
overflowing a 32-bit value, but I don't have any 32-bit machines handy
to test
Based on what you've told us it looks like maybe you're running into a
btrfs bug of some kind.
We're working on getting the most recent version of btrfs onto some
nodes here to check it out as we've been hearing about sync issues
with btrfs for a few days now, but we haven't managed it yet.
1001 - 1100 of 1146 matches
Mail list logo