]> git.apps.os.sepia.ceph.com Git - ceph.git/log
ceph.git
12 years agofusetrace_ll.cc: reduce scope of 'int err' in main()
Danny Al-Gaaf [Mon, 11 Mar 2013 11:11:06 +0000 (12:11 +0100)]
fusetrace_ll.cc: reduce scope of 'int err' in main()

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
12 years agomds/Locker.cc: fix warning about 'Possible null pointer dereference'
Danny Al-Gaaf [Fri, 1 Mar 2013 19:12:03 +0000 (20:12 +0100)]
mds/Locker.cc: fix warning about 'Possible null pointer dereference'

Fix following warning from cppcheck:
 [src/mds/Locker.cc:2255] -> [src/mds/Locker.cc:2258]: (error)
   Possible null pointer dereference: in - otherwise it is redundant
   to check it against null.

Since head_in used for call pick_inode_snap() is already valid,
there is no need to check if 'CInode *in' is not NULL. Remove
not needed check.

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
12 years agomds/Locker.cc: use static_cast instead of C-Style cast
Danny Al-Gaaf [Fri, 1 Mar 2013 18:55:05 +0000 (19:55 +0100)]
mds/Locker.cc: use static_cast instead of C-Style cast

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
12 years agoMerge remote-tracking branch 'gh/next'
Sage Weil [Sat, 9 Mar 2013 02:38:32 +0000 (18:38 -0800)]
Merge remote-tracking branch 'gh/next'

12 years agoceph_common.sh: add warning if 'host' contains dots
Dan Mick [Fri, 8 Mar 2013 23:18:54 +0000 (15:18 -0800)]
ceph_common.sh: add warning if 'host' contains dots

This is a common error and there's no reason the script can't
at least tell you it's a really bad idea.  One might argue it
could even successfully proactively truncate the host parameter
at the first dot, but that's a little controlling, perhaps.

Signed-off-by: Dan Mick <dan.mick@inktank.com>
Reviewed-by: Josh Durgin <josh.durgin@inktank.com>
12 years agoRevert "Update conf.py"
Dan Mick [Fri, 8 Mar 2013 23:46:34 +0000 (15:46 -0800)]
Revert "Update conf.py"

This reverts commit 700899206224436255522d8469612e6603fe3116.
Apparently github's edit lets you commit directly to the repo.
Oops.

12 years agoUpdate conf.py
Dan Mick [Fri, 8 Mar 2013 23:44:28 +0000 (15:44 -0800)]
Update conf.py

12 years agoMerge pull request #96 from concubidated/master
Sage Weil [Fri, 8 Mar 2013 21:49:04 +0000 (13:49 -0800)]
Merge pull request #96 from concubidated/master

Fix output of command 'ceph osd tree --format=json'

Reviewed-by: Sage Weil <sage@inktank.com>
12 years agodoc: Added a new network configuration reference document.
John Wilkins [Fri, 8 Mar 2013 21:45:04 +0000 (13:45 -0800)]
doc: Added a new network configuration reference document.

Signed-off-by: John Wilkins <john.wilkins@inktank.com>
12 years agoFix output of 'ceph osd tree --format=json' 96/head
Concubidated [Fri, 8 Mar 2013 21:44:39 +0000 (13:44 -0800)]
Fix output of 'ceph osd tree --format=json'

Signed-off-by: Tyler Brekke <tyler.brekke@inktank.com>
12 years agodoc: Added network settings. Reorganized index.
John Wilkins [Fri, 8 Mar 2013 21:44:31 +0000 (13:44 -0800)]
doc: Added network settings. Reorganized index.

Signed-off-by: John Wilkins <john.wilkins@inktank.com>
12 years agodoc: Excised network settings.
John Wilkins [Fri, 8 Mar 2013 21:43:44 +0000 (13:43 -0800)]
doc: Excised network settings.

Signed-off-by: John Wilkins <john.wilkins@inktank.com>
12 years agodoc: Format edits. Excised much of network discussion. Added reference to network...
John Wilkins [Fri, 8 Mar 2013 21:43:12 +0000 (13:43 -0800)]
doc: Format edits. Excised much of network discussion. Added reference to network guide.

Signed-off-by: John Wilkins <john.wilkins@inktank.com>
12 years agoMerge branch 'wip-osd-map' into next
Samuel Just [Fri, 8 Mar 2013 20:44:42 +0000 (12:44 -0800)]
Merge branch 'wip-osd-map' into next

Fixes: 4369
Backport: bobtail
Reviewed-by: Samuel Just <sam.just@inktank.com>
12 years agoosd: mark down connections from old peers
Sage Weil [Fri, 8 Mar 2013 16:56:44 +0000 (08:56 -0800)]
osd: mark down connections from old peers

Close out any connection with an old peer.  This avoids a race like:

- peer marked down
- we get map, mark down the con
- they reconnect and try to send us some stuff
- we share our map to tell them they are old and dead, but leave the con
  open
...
- peer marks itself up a few times, eventually reuses the same port
- sends messages on their fresh con
- we discard because of our old con

This could cause a tight reconnect loop, but it is better than wrong
behavior.

Other possible fixes:
 - make addr nonce truly unique (augment pid in nonce)
 - make a smarter 'disposable' msgr state (bleh)

Signed-off-by: Sage Weil <sage@inktank.com>
12 years agoosd/PG: rename require_same_or_newer_map -> is_same_or_newer_map
Sage Weil [Fri, 8 Mar 2013 16:53:40 +0000 (08:53 -0800)]
osd/PG: rename require_same_or_newer_map -> is_same_or_newer_map

This avoids confusion with the OSD method of the same name, and better
matches what the function tests (and does not do).

Signed-off-by: Sage Weil <sage@inktank.com>
12 years agomon/Paxos: make old clock skew message more/less suggestive
Sage Weil [Fri, 8 Mar 2013 18:43:21 +0000 (10:43 -0800)]
mon/Paxos: make old clock skew message more/less suggestive

There is a separate clock skew check now, and more often than not this
suggests laggy messages, and not necessarily skew.

Signed-off-by: Sage Weil <sage@inktank.com>
12 years agoMerge pull request #92 from jaharkes/content_length
Yehuda Sadeh [Fri, 8 Mar 2013 17:45:44 +0000 (09:45 -0800)]
Merge pull request #92 from jaharkes/content_length

Properly format Content-Length: header on 32-bit systems.

Reviewed-by: Yehuda Sadeh <yehuda@inktank.com>
12 years agoProperly format Content-Length: header on 32-bit systems. 92/head
Jan Harkes [Thu, 7 Mar 2013 21:07:26 +0000 (16:07 -0500)]
Properly format Content-Length: header on 32-bit systems.

- Promote len argument in dump_content_length to uint64_t.
- Make sure there is sufficient scratch space to format string.
- Use PRIu64 macro for formatting.

Signed-off-by: Jan Harkes <jaharkes@cs.cmu.edu>
12 years agorgw: set attrs on various list bucket xml results (swift)
Yehuda Sadeh [Fri, 22 Feb 2013 23:04:37 +0000 (15:04 -0800)]
rgw: set attrs on various list bucket xml results (swift)

Fixes: #4247
The list buckets operation was missing some attrs on the different
xml result entities. This fixes it.

Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
12 years agoformatter: add the ability to dump attrs in xml entities
Yehuda Sadeh [Fri, 22 Feb 2013 23:02:02 +0000 (15:02 -0800)]
formatter: add the ability to dump attrs in xml entities

xml entities may have attrs assigned to them. Add the ability
to set them. A usage example:

formatter->open_array_section_with_attrs("container",
     FormatterAttrs("name", "foo", NULL));

This will generate the following xml entity:
<container name="foo">

Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
Reviewed-by: Greg Farnum <greg@inktank.com>
12 years agorgw: don't iterate through all objects when in namespace
Yehuda Sadeh [Thu, 7 Mar 2013 03:32:21 +0000 (19:32 -0800)]
rgw: don't iterate through all objects when in namespace

Fixes: #4363
Backport: argonaut, bobtail
When listing objects in namespace don't iterate through all the
objects, only go though the ones that starts with the namespace
prefix

Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
12 years agoMerge pull request #93 from jaharkes/range_requests
Yehuda Sadeh [Fri, 8 Mar 2013 14:47:00 +0000 (06:47 -0800)]
Merge pull request #93 from jaharkes/range_requests

Avoid sending a success response when an error occurs.

Reviewed-by: Yehuda Sadeh <yehuda@inktank.com>
12 years agoMerge branch 'wip-2733'
David Zafman [Fri, 8 Mar 2013 04:02:04 +0000 (20:02 -0800)]
Merge branch 'wip-2733'

12 years agoAdd filestore_replica_fadvise config option default true
David Zafman [Fri, 8 Mar 2013 02:02:07 +0000 (18:02 -0800)]
Add filestore_replica_fadvise config option default true

Signed-off-by: David Zafman <david.zafman@inktank.com>
Reviewed-by: Sam Just <sam.just@inktank.com>
12 years agoosd: fadvise replica data don't use
David Zafman [Fri, 8 Mar 2013 03:36:07 +0000 (19:36 -0800)]
osd: fadvise replica data don't use

Add transaction flag indicating if op is at a replica
After write has been sync'ed use posix_fadvise() to clear kernel cache

Feature: #2733

Signed-off-by: David Zafman <david.zafman@inktank.com>
Reviewed-by: Sam Just <sam.just@inktank.com>
12 years agoAvoid sending a success response when an error occurs. 93/head
Jan Harkes [Fri, 8 Mar 2013 03:10:42 +0000 (22:10 -0500)]
Avoid sending a success response when an error occurs.

Functions called from RGWGetObj_ObjStore_S3::send_response_data may
change the value of the non-local variable 'ret'. But the response
relies on a local 'req_state' which copies ret at the start of the
function.

Right now none of the called functions actually changes ret so the
problem doesn't trigger, but to avoid future breakage it is safer
to not rely on the (early) copy of the ret variable.

Signed-off-by: Jan Harkes <jaharkes@cs.cmu.edu>
12 years agoFix FileStore to notice if filestore_flusher config is changed
David Zafman [Fri, 8 Mar 2013 01:05:20 +0000 (17:05 -0800)]
Fix FileStore to notice if filestore_flusher config is changed

Signed-off-by: David Zafman <david.zafman@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>
12 years agoMerge branch 'wip-traceless'
Sage Weil [Fri, 8 Mar 2013 01:52:26 +0000 (17:52 -0800)]
Merge branch 'wip-traceless'

Reviewed-by: Greg Farnum <greg@inktank.com>
12 years agoclient: move annoying traceless handling into verify_reply_trace()
Sage Weil [Fri, 8 Mar 2013 00:31:29 +0000 (16:31 -0800)]
client: move annoying traceless handling into verify_reply_trace()

Signed-off-by: Sage Weil <sage@inktank.com>
12 years agoclient: debug async cache invalidation
Sage Weil [Wed, 6 Mar 2013 23:51:48 +0000 (15:51 -0800)]
client: debug async cache invalidation

Signed-off-by: Sage Weil <sage@inktank.com>
12 years agomds: pass created_ino back to client on replayed requests
Sage Weil [Wed, 6 Mar 2013 03:12:21 +0000 (19:12 -0800)]
mds: pass created_ino back to client on replayed requests

After an MDS restart, the client will resend uncommitted requests.  Use the
information we now have in the session_info_t to pass the created ino
back via the extra_bl payload in the reply.

Fixes: #4034
Signed-off-by: Sage Weil <sage@inktank.com>
12 years agomds: track created inos with completed request ids in session info
Sage Weil [Fri, 8 Mar 2013 00:33:12 +0000 (16:33 -0800)]
mds: track created inos with completed request ids in session info

Along with each session completed request (tid), also track the created
ino (if any).  This will be used to pass back to the client when they
replay requests in the next patch.

Do not bother making this a backward compatible encoding.  The only
benefit is to allow ceph-mds code to run and then old mds code to run
after it.  Given all of the other incompat changes we *just* made, this
is highly unlikely and not worth the code clutter.  If we had spanned
more releases or a stable release the story would be different.

While we are here, inline the second add_completed_request() method variant
since there is only a single caller and having it overloaded somewhat
obscures what is going on.  This also avoids a duplicate lookup in the
session map in the have_session() check and then in the (old) helper.

Signed-off-by: Sage Weil <sage@inktank.com>
12 years agoclient: debug created ino
Sage Weil [Wed, 6 Mar 2013 02:30:05 +0000 (18:30 -0800)]
client: debug created ino

Signed-off-by: Sage Weil <sage@inktank.com>
12 years agoclient: unlink old_dentry on traceless rename reply
Sage Weil [Mon, 4 Mar 2013 23:47:06 +0000 (15:47 -0800)]
client: unlink old_dentry on traceless rename reply

Signed-off-by: Sage Weil <sage@inktank.com>
12 years agoclient: force lookup on traceless reply
Sage Weil [Mon, 4 Mar 2013 23:46:49 +0000 (15:46 -0800)]
client: force lookup on traceless reply

Call _do_lookup directly so that we avoid hitting the cache

Signed-off-by: Sage Weil <sage@inktank.com>
12 years agoclient: make _do_lookup take a const string
Sage Weil [Mon, 4 Mar 2013 23:45:51 +0000 (15:45 -0800)]
client: make _do_lookup take a const string

Signed-off-by: Sage Weil <sage@inktank.com>
12 years agomds: do not do traceless reply for open O_TRUNC requests
Sage Weil [Sat, 2 Mar 2013 00:28:25 +0000 (16:28 -0800)]
mds: do not do traceless reply for open O_TRUNC requests

Even though these are "may_write" and may be a mutation, the MDS should not
fake out a traceless reply.  In a real failure, it looks like:

 - mds process request (mutation)
 - mds fails
 - client resends request
 - mds notes that it is O_TRUNC, and already committed, and proceeds with
   a regular open

And that is all fine.

But, that means that if we are trying to simulate this behavior without a
failure, then we should never do a traceless reply on an O_TRUNC request,
because the client will never see such a reply--it'll get a full open
request response instead.

Signed-off-by: Sage Weil <sage@inktank.com>
12 years agoclient: handle fh cleanup in ll_create() failure path
Sage Weil [Thu, 7 Mar 2013 23:22:26 +0000 (15:22 -0800)]
client: handle fh cleanup in ll_create() failure path

Release the fh if we fail the permissions check.

Signed-off-by: Sage Weil <sage@inktank.com>
12 years agoclient: root can write to any file
Sage Weil [Fri, 1 Mar 2013 23:57:40 +0000 (15:57 -0800)]
client: root can write to any file

Signed-off-by: Sage Weil <sage@inktank.com>
12 years agoclient: clear I_COMPLETE on traceless reply for dentry request
Sage Weil [Fri, 1 Mar 2013 06:32:30 +0000 (22:32 -0800)]
client: clear I_COMPLETE on traceless reply for dentry request

If a request is against a dentry, and we get a traceless reply, clear
the directory I_COMPLETE flag on the parent directory because we can no
longer trust that our cache is complete.

It is possible we could do something a bit more intelligent here, but it
is not trivial because of racing requests, and traceless replies are
rare, so it's not worth the effort.

Signed-off-by: Sage Weil <sage@inktank.com>
12 years agoclient: pass ptarget from link, create, mkdir, symlink, setattr
Sage Weil [Fri, 1 Mar 2013 02:08:51 +0000 (18:08 -0800)]
client: pass ptarget from link, create, mkdir, symlink, setattr

Use the new make_request functionality to ensure that we properly handle
the re-lookup/getattr when we get a traceless reply back from the MDS.

Note that we have to do this on setattr() because ll_getattr() expects to
return the updated inode metadata to the caller.

Signed-off-by: Sage Weil <sage@inktank.com>
12 years agoclient: handle traceless replies from make_request()
Sage Weil [Fri, 1 Mar 2013 02:07:33 +0000 (18:07 -0800)]
client: handle traceless replies from make_request()

Modify the generic make_request wrapper to retry lookups or getattrs on
requests when the ptarget pointer is specified but there is no trace in
the reply.

Refactor the _create() method to use this, effectively moving all of the
extra_bl code processing into make_request where it is generically
useful.

Signed-off-by: Sage Weil <sage@inktank.com>
12 years agoosd: increate default pg log size from 1000 -> 3000
Sage Weil [Thu, 7 Mar 2013 18:31:27 +0000 (10:31 -0800)]
osd: increate default pg log size from 1000 -> 3000

This reduces the probability that we will fail to detect a dup op.  See
#4368.

Signed-off-by: Sage Weil <sage@inktank.com>
12 years agoMerge remote-tracking branch 'gh/wip-log-max' into next
Sage Weil [Thu, 7 Mar 2013 17:29:44 +0000 (09:29 -0800)]
Merge remote-tracking branch 'gh/wip-log-max' into next

Reviewed-by: Sage Weil <sage@inktank.com>
12 years agovstart.sh: osd debug op order = true
Sage Weil [Thu, 7 Mar 2013 05:34:36 +0000 (21:34 -0800)]
vstart.sh: osd debug op order = true

Signed-off-by: Sage Weil <sage@inktank.com>
12 years agoconfig: note which options are overridden by common_preinit()
Josh Durgin [Thu, 7 Mar 2013 01:49:51 +0000 (17:49 -0800)]
config: note which options are overridden by common_preinit()

Defaults for these differ based on the context in which they're used.

Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
12 years agocommon: reduce default in-memory logs for non-daemons
Josh Durgin [Thu, 7 Mar 2013 01:42:03 +0000 (17:42 -0800)]
common: reduce default in-memory logs for non-daemons

The default of 100000 can result in hundreds of MBs of extra memory
used. This was most obvious when using librbd with caching enabled,
since there was a dout(0) accidentally left in the ObjectCacher.

refs: #4352
backport: bobtail
Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
12 years agoinit-ceph: fix run dir
Sage Weil [Thu, 7 Mar 2013 01:09:51 +0000 (17:09 -0800)]
init-ceph: fix run dir

Signed-off-by: Sage Weil <sage@inktank.com>
12 years agoosd: add ctor for clone_info
Sage Weil [Wed, 6 Mar 2013 18:21:59 +0000 (10:21 -0800)]
osd: add ctor for clone_info

Signed-off-by: Sage Weil <sage@inktank.com>
(cherry picked from commit 439d0e334d7d0246806e5b7c6702b8198f9bed71)

12 years agoMissed adding rados_types.hpp to package
David Zafman [Wed, 6 Mar 2013 22:33:12 +0000 (14:33 -0800)]
Missed adding rados_types.hpp to package

Caused by 3bd48cbbadb7908dd833ccde75359f085828fc5c
feature 4207 implementation

Signed-off-by: David Zafman <david.zafman@inktank.com>
Reviewed-by: Gary Lowell <gary.lowell@inktank.com>
12 years agoObjectCacher: fix debug log level in split
Josh Durgin [Thu, 28 Feb 2013 20:13:45 +0000 (12:13 -0800)]
ObjectCacher: fix debug log level in split

Level 0 should never be used for this kind of debugging.

Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
Reviewed-by: Dan Mick <dan.mick@inktank.com>
12 years agoosd: add ctor for clone_info
Sage Weil [Wed, 6 Mar 2013 18:21:59 +0000 (10:21 -0800)]
osd: add ctor for clone_info

Signed-off-by: Sage Weil <sage@inktank.com>
12 years agoMerge remote-tracking branch 'gh/wip-expose-topo'
Sage Weil [Wed, 6 Mar 2013 16:11:26 +0000 (08:11 -0800)]
Merge remote-tracking branch 'gh/wip-expose-topo'

Reviewed-by: Sage Weil <sage@inktank.com>
12 years agoMerge pull request #90 from grosskur/fix-debian-libsnappy
Sage Weil [Wed, 6 Mar 2013 15:54:13 +0000 (07:54 -0800)]
Merge pull request #90 from grosskur/fix-debian-libsnappy

debian: require libsnappy-dev for ceph

12 years agodebian: require libsnappy-dev for ceph 90/head
Alan Grosskurth [Wed, 6 Mar 2013 10:21:12 +0000 (02:21 -0800)]
debian: require libsnappy-dev for ceph

Debian builds are currently broken without this requirement.

12 years agolibcephfs: return osd location from crush map
Noah Watkins [Wed, 6 Mar 2013 02:00:26 +0000 (18:00 -0800)]
libcephfs: return osd location from crush map

Adds ceph_get_osd_crush_location that returns a list of (type,name)
pairs for devices that exist between a given osd and the root of the
bucket hierarchy.

Signed-off-by: Noah Watkins <noahwatkins@gmail.com>
12 years agorelease-process.rst: Fix typos
Gary Lowell [Wed, 6 Mar 2013 06:08:15 +0000 (22:08 -0800)]
release-process.rst:  Fix typos

Signed-off-by: Gary Lowell <gary.lowell@inktank.com>
12 years agoclient: expose extent/osd mapping
Noah Watkins [Fri, 22 Feb 2013 23:46:02 +0000 (15:46 -0800)]
client: expose extent/osd mapping

Signed-off-by: Noah Watkins <noahwatkins@gmail.com>
12 years agoMerge branch 'wip-prepare'
Sage Weil [Tue, 5 Mar 2013 21:33:05 +0000 (13:33 -0800)]
Merge branch 'wip-prepare'

Reviewed-by: Josh Durgin <josh.durgin@inktank.com>
Reviewed-by: Alexandre Marangone <alexandre.marangone@inktank.com>
Tested-by: Tamil Muthamizhan <tamil.muthamizhan@inktank.com>
12 years agoceph-disk-prepare: move in-use checks to the top, before zap
Sage Weil [Tue, 5 Mar 2013 21:08:26 +0000 (13:08 -0800)]
ceph-disk-prepare: move in-use checks to the top, before zap

Move the in-use checks to the very top, before we (say) zap!

Signed-off-by: Sage Weil <sage@inktank.com>
12 years agodoc/release-notes: v0.58
Sage Weil [Tue, 5 Mar 2013 19:02:19 +0000 (11:02 -0800)]
doc/release-notes: v0.58

Signed-off-by: Sage Weil <sage@inktank.com>
12 years agoUpdate doc/radosgw/s3/python.rst
carsonoid [Thu, 21 Feb 2013 22:49:17 +0000 (15:49 -0700)]
Update doc/radosgw/s3/python.rst

Note about creating conn object for non-ssl implementations.

12 years agoMerge branch 'wip-4207'
David Zafman [Tue, 5 Mar 2013 18:25:49 +0000 (10:25 -0800)]
Merge branch 'wip-4207'

Feature: #4207: osd/librados: add ops to list snaps for an object

Signed-off-by: David Zafman <david.zafman@inktank.com>
Reported-by: Sam Just <sam.just@inktank.com>
12 years agoMerge branch 'master' of https://github.com/ceph/ceph
Gary Lowell [Tue, 5 Mar 2013 17:06:24 +0000 (09:06 -0800)]
Merge branch 'master' of https://github.com/ceph/ceph

12 years agoMerge branch 'next'
Gary Lowell [Tue, 5 Mar 2013 17:05:21 +0000 (09:05 -0800)]
Merge branch 'next'

12 years agoAdd list_snaps() calls to snapshots test cases
David Zafman [Sat, 2 Mar 2013 02:27:03 +0000 (18:27 -0800)]
Add list_snaps() calls to snapshots test cases

Signed-off-by: David Zafman <david.zafman@inktank.com>
12 years agoAdd rados listsnaps command
David Zafman [Sat, 23 Feb 2013 04:36:28 +0000 (20:36 -0800)]
Add rados listsnaps command

Signed-off-by: David Zafman <david.zafman@inktank.com>
12 years agoosd/librados: add op to list clones/snaps for an object
David Zafman [Fri, 22 Feb 2013 23:15:06 +0000 (15:15 -0800)]
osd/librados: add op to list clones/snaps for an object

Returning snap_set_t with clone info
and snapshots in ascending order
Add clones with snapshots to obj_list_snap_response_t
New rados_types.hpp with snap_set_t/clone_info_t
Move snap_t to rados_types.hpp
Add generate_test_instances() and TYPE() to encoding/types.h

Feature: #4207

Signed-off-by: David Zafman <david.zafman@inktank.com>
12 years agoRemove unused SnapContext in librados.hpp
David Zafman [Thu, 28 Feb 2013 01:49:17 +0000 (17:49 -0800)]
Remove unused SnapContext in librados.hpp

Signed-off-by: David Zafman <david.zafman@inktank.com>
12 years agoosd: Improve snapshot test
David Zafman [Wed, 27 Feb 2013 04:54:48 +0000 (20:54 -0800)]
osd: Improve snapshot test

In SelfManagedSnapRollbackPP add some overlapping writes
New SelfManagedSnapOverlap creates overlapping writes

Signed-off-by: David Zafman <david.zafman@inktank.com>
12 years agoosd: correct comment
David Zafman [Wed, 27 Feb 2013 01:09:14 +0000 (17:09 -0800)]
osd: correct comment

SnapSet snaps are in descending order

Signed-off-by: David Zafman <david.zafman@inktank.com>
12 years agoceph-disk-prepare: verify device is not in use by device-mapper
Sage Weil [Tue, 5 Mar 2013 04:57:52 +0000 (20:57 -0800)]
ceph-disk-prepare: verify device is not in use by device-mapper

Be nice and tell the user which devices/mappings are consuming the device,
too.

Signed-off-by: Sage Weil <sage@inktank.com>
12 years agoceph-disk-prepare: verify device is not mounted before using
Sage Weil [Tue, 5 Mar 2013 00:38:57 +0000 (16:38 -0800)]
ceph-disk-prepare: verify device is not mounted before using

Make sure the data and/or journal device(s) are not in use (mounted)
before using them.  Make room for additional "in-use" checks in the future.

Closes: #3256
Signed-off-by: Sage Weil <sage@inktank.com>
12 years agodebian: require >= python2.6 for ceph as well
Josh Durgin [Tue, 5 Mar 2013 00:47:49 +0000 (16:47 -0800)]
debian: require >= python2.6 for ceph as well

ceph-disk-prepare and ceph-disk-activate use the with statement,
str.format, and possibly other new features from python 2.6.

Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
12 years agoceph-disk-prepare: clean up stupid check for a digit
Sage Weil [Tue, 5 Mar 2013 00:29:57 +0000 (16:29 -0800)]
ceph-disk-prepare: clean up stupid check for a digit

Signed-off-by: Sage Weil <sage@inktank.com>
12 years agoceph-disk-prepare: use os.path.realpath()
Sage Weil [Tue, 5 Mar 2013 00:08:15 +0000 (16:08 -0800)]
ceph-disk-prepare: use os.path.realpath()

My janky symlink resolution is broken in various ways.

Signed-off-by: Sage Weil <sage@inktank.com>
12 years agov0.58 v0.58
Gary Lowell [Mon, 4 Mar 2013 23:18:24 +0000 (15:18 -0800)]
v0.58

12 years agoAdd X-Python-Version >=2.6 to debian control file.
Filippos Giannakos [Wed, 27 Feb 2013 11:56:08 +0000 (13:56 +0200)]
Add X-Python-Version >=2.6 to debian control file.

python-ceph complains when installed to debian squeeze about the 'with'
statement. Apparently installation tries to install the python-ceph package for
python 2.5, which does not support the 'with' statement natively.

Signed-off-by: Filippos Giannakos <philipgian@grnet.gr>
12 years agoPG,ReplicatedPG: use pg_has_reset_since to discard old async events
Samuel Just [Mon, 25 Feb 2013 21:28:31 +0000 (13:28 -0800)]
PG,ReplicatedPG: use pg_has_reset_since to discard old async events

Local async events are obsolete if the pg is deleting or if the
epoch at which they were created is prior to last_peering_reset.

Signed-off-by: Samuel Just <sam.just@inktank.com>
(cherry picked from commit 62999ea2d362b32dfc86a15dcb03b06fb5cb0572)

12 years agoPG::build_scrub_map: detect race with peering via last_peering_reset
Samuel Just [Mon, 25 Feb 2013 20:40:06 +0000 (12:40 -0800)]
PG::build_scrub_map: detect race with peering via last_peering_reset

Signed-off-by: Samuel Just <sam.just@inktank.com>
(cherry picked from commit 67225339dc3d62d7fe5a32eec65d51e53e8d35bb)

12 years agoReplicatedPG::C_OSD_CommittedPushedObject: use intrusive_ptr for pg
Samuel Just [Mon, 25 Feb 2013 20:36:29 +0000 (12:36 -0800)]
ReplicatedPG::C_OSD_CommittedPushedObject: use intrusive_ptr for pg

Signed-off-by: Samuel Just <sam.just@inktank.com>
(cherry picked from commit 04ee8f478bbd587a711d0668c471cfc5c1cab06c)

12 years agoReplicatedPG::C_OSD_CommittedPushedObject take epoch submitted
Samuel Just [Mon, 25 Feb 2013 20:35:26 +0000 (12:35 -0800)]
ReplicatedPG::C_OSD_CommittedPushedObject take epoch submitted

What we really care about is that the epoch in which the Context
was submitted is at complete() time >= last_peering_reset.

Signed-off-by: Samuel Just <sam.just@inktank.com>
(cherry picked from commit a01dea6af9aacf0614570ebb5fa161d9dde9b6b6)

12 years agoinit-ceph: fix --restart option
Sage Weil [Mon, 4 Mar 2013 20:34:40 +0000 (12:34 -0800)]
init-ceph: fix --restart option

Reported-by: Bryan K. Wright <bkw1a@ayesha.phys.virginia.edu>
Signed-off-by: Sage Weil <sage@inktank.com>
12 years agodoc: Updated year in footer.
John Wilkins [Mon, 4 Mar 2013 19:48:14 +0000 (11:48 -0800)]
doc: Updated year in footer.

Signed-off-by: John Wilkins <john.wilkins@inktank.com>
12 years agoMerge branch 'master' of https://github.com/ceph/ceph
John Wilkins [Mon, 4 Mar 2013 18:59:50 +0000 (10:59 -0800)]
Merge branch 'master' of https://github.com/ceph/ceph

12 years agodoc: Added instructions for ensuring ceph config file and keyring prior to mounting...
John Wilkins [Mon, 4 Mar 2013 18:59:18 +0000 (10:59 -0800)]
doc: Added instructions for ensuring ceph config file and keyring prior to mounting ceph fuse.

fixes: #3796

Signed-off-by: John Wilkins <john.wilkins@inktank.com>
12 years agoMerge pull request #89 from Elbandi/master
Sage Weil [Mon, 4 Mar 2013 18:33:15 +0000 (10:33 -0800)]
Merge pull request #89 from Elbandi/master

remove unused variable from makefile

12 years agojounal: disable aio
Sage Weil [Mon, 4 Mar 2013 18:08:49 +0000 (10:08 -0800)]
jounal: disable aio

There is a deadlock issue in the aio code, see #4079.  Disable for the time
being.

Signed-off-by: Sage Weil <sage@inktank.com>
12 years agoMerge remote-tracking branch 'gh/wip-3484'
Sage Weil [Mon, 4 Mar 2013 17:18:28 +0000 (09:18 -0800)]
Merge remote-tracking branch 'gh/wip-3484'

Reviewed-by: Sage Weil <sage@inktank.com>
12 years agoBuild: no longer need to use LEVELDB_INCLUDE 89/head
Andras Elso [Fri, 1 Mar 2013 15:34:01 +0000 (16:34 +0100)]
Build: no longer need to use LEVELDB_INCLUDE

Signed-off-by: Andras Elso <elso.andras@gmail.com>
12 years agotools: ceph: add 'ceph df' usage
Joao Eduardo Luis [Sat, 2 Mar 2013 18:36:37 +0000 (18:36 +0000)]
tools: ceph: add 'ceph df' usage

Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
12 years agomon: Monitor: 'ceph df'
Joao Eduardo Luis [Fri, 1 Mar 2013 16:32:24 +0000 (16:32 +0000)]
mon: Monitor: 'ceph df'

Fixes: #3484
Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
12 years agoREADME: Add libsnappy-dev to build requirements.
Gary Lowell [Sat, 2 Mar 2013 07:32:54 +0000 (23:32 -0800)]
README:  Add libsnappy-dev to build requirements.

Signed-off-by: Gary Lowell <gary.lowell@inktank.com>
12 years agodoc: missing blank line in radosgw/config-ref
Dan Mick [Sat, 2 Mar 2013 03:53:33 +0000 (19:53 -0800)]
doc: missing blank line in radosgw/config-ref

12 years agoMerge pull request #85 from dalgaaf/wip-da-spec-1
Gary Lowell [Sat, 2 Mar 2013 03:09:04 +0000 (19:09 -0800)]
Merge pull request #85 from dalgaaf/wip-da-spec-1

Fix package requirements to get sgdisk and merge some changes back from SUSE spec file.

Reviewed-by: Gary Lowell <gary.lowell@inktank.com>
12 years agoMerge pull request #88 from nmtadam/master
Sage Weil [Sat, 2 Mar 2013 01:50:00 +0000 (17:50 -0800)]
Merge pull request #88 from nmtadam/master

librados.hpp typo

12 years agolibrados.hpp: replace paris with pairs 88/head
Adam Manzanares [Fri, 1 Mar 2013 23:49:26 +0000 (15:49 -0800)]
librados.hpp: replace paris with pairs

Signed-off-by: Adam Manzanares <nmtadam@gmail.com>
Working on a lua-rados library I noticed that paris was used
in the place of pairs in some comments.

12 years agodoc: Instruction should have been noout, not nodown.
John Wilkins [Fri, 1 Mar 2013 23:09:31 +0000 (15:09 -0800)]
doc: Instruction should have been noout, not nodown.

Signed-off-by: John Wilkins <john.wilkins@inktank.com>
12 years agodoc: Fixed OSD pool config settings. Should be in [global], not [osd].
John Wilkins [Fri, 1 Mar 2013 19:35:47 +0000 (11:35 -0800)]
doc: Fixed OSD pool config settings. Should be in [global], not [osd].

Signed-off-by: John Wilkins <john.wilkins@inktank.com>