RepoMirrors/ceph
1
0
Fork 0
mirror of https://github.com/ceph/ceph synced 2025-02-07 19:03:18 +00:00
Code Issues Packages Projects Releases Wiki Activity
50,519 Commits 215 Branches 395 Tags 1.7 GiB
b62add00a0
Commit Graph

50519 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Sage Weil
fafb14212d fix init of new fields in MCLientCaps 2016-03-01 11:18:17 -05:00
Sage Weil
7f47cd76d2 mds/MDSMap: add file layout v2 incompat feature 
The existing MDSMonitor machinations already add this to the MDSMap to
boot out old MDS's.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 11:18:17 -05:00
Sage Weil
00850a67ca mds: deny access to directories with pool_ns layouts if feature is missing 
If the client feature is missing, given them EIO.  Change check_access to
return an error code to distinguish this from EACCES.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 11:18:17 -05:00
Sage Weil
9c7669b95a mon/MDSMonitor: prevent pool 0 from being used as a data pool 
Pool 0 means no change or default in the legacy ceph_file_layout in the
layout ioctl and file create arguments. Prevent it from being used to avoid
putting users in an awkward situation later.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 11:18:17 -05:00
Yan, Zheng
a2d42d5572 mds: don't break compatibility of MClientCaps/MClientReply 
Signed-off-by: Yan, Zheng <zyan@redhat.com>
 2016-03-01 11:18:17 -05:00
Sage Weil
efb4a83f77 fs_types: file_layout_t: convert pool -1 (undefined) to 0 in legacy encoding 
Old code assumes that fl_pg_pool == 0 means the pool is not defined, while
file_layout_t uses -1.  Translate between the two.

Note that this means a valid file_layout_t with pool_id == 0 cannot be
accurately translated to a legacy file_layout_t.  That is somewhat
unavoidable, and should not be a problem since real clusters create 'rbd'
as pool 0 and it does not use any file layouts.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 11:18:17 -05:00
Sage Weil
c31e94d3b5 qa/workunits/fs/misc/layout_vxattrs: test layout.pool_namespace 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 11:18:17 -05:00
Sage Weil
e0ad651744 client: expose layout.pool_namespace vxattr 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 11:18:16 -05:00
Sage Weil
386d7c6b02 mds: allow adjustment of layout.pool_namespace vxattr 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 11:18:16 -05:00
Sage Weil
99f1f6478d struct ceph_file_layout -> file_layout_t 
- drop the global
- do not memset!
- encode with features
- field names are different
- use get_period() method where appropriate
- fix is layout empty checks

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 11:18:16 -05:00
Sage Weil
2752eeefc3 mds: make inode_t, old_inode_t featureful 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 11:17:41 -05:00
Sage Weil
e750f48c03 messages/MClientCaps: avoid fixed struct for encoding 
At least for new clients.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 11:17:41 -05:00
Sage Weil
ff606bf4c4 mds: encode inode reply information directly 
Drop the ceph_mds_reply_inode struct and encode directly.  Note
that we already decode directly to InodeStat in MClientReply.h.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 11:17:00 -05:00
Sage Weil
0b1f7f59b6 messages/MClientReply: stop using ceph_mds_reply_inode struct to unpack 
Decode directly.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 11:17:00 -05:00
Sage Weil
7dcf39f7ca mds: pass features to various CInode encoders and helpers 
We'll need this shortly.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 11:17:00 -05:00
Sage Weil
826e8d01dc mds/MDSMap: cache up_features; fix logic 
& these, not | them.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 11:17:00 -05:00
Sage Weil
162ca5bcf9 libradosstriper: move default stripe parameters inline 
The extern is going away.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 11:17:00 -05:00
Sage Weil
7b961ff129 include/encoding: featureful encoder for list of shared_ptrs 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 11:16:59 -05:00
Sage Weil
bdae5bf37c include/compact_map: featureful encoders, too 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 11:16:59 -05:00
Sage Weil
4cf4f8dd13 ceph-dencoder: TYPE_FEATUREFUL_NOCOPY 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 11:16:59 -05:00
Sage Weil
358cce1e7f mds: pass features to LogEvents, EMetaBlob 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 11:16:59 -05:00
Sage Weil
e08dd70d78 mds,mon: include features in beacon messages, MDSMap 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 11:16:59 -05:00
Sage Weil
4ad8f7254a mds/MDSMap: add features to MDSMap's mds_info_t 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 11:16:59 -05:00
Sage Weil
982208135c add fs_types.h, with file_layout_t 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 11:16:59 -05:00
Sage Weil
0cd1b94133 declare Formatter for everyone 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 11:16:59 -05:00
Sage Weil
58a00e01f5 qa/workunits/fs/misc/layout_vxattr.sh: detect data pool name 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 11:16:58 -05:00
branto1
cc7f9ec63c Merge pull request #7823 from SUSE/wip-14906 
packaging: add lsb-release dependency for SUSE and Debian

Reviewed-by: David Disseldorp <ddiss@suse.de>
Reviewed-by: Ken Dreyer <kdreyer@redhat.com>
Reviewed-by: Boris Ranto <branto@redhat.com>
 2016-03-01 16:35:21 +01:00
Radoslaw Zarzynski
533b617739 rgw: send proper ETag value during GET on DLO/SLO through S3 API. 
Signed-off-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
 2016-03-01 16:26:33 +01:00
Sage Weil
efd7625f38 Merge pull request #7837 from rldleblanc/wpq-documentation 
doc: osd-config Add Configuration Options for op queue.
 2016-03-01 09:27:21 -05:00
Sage Weil
805e37cf8b Merge pull request #7851 from jupiturliu/obj-bencher-bug 
rados: fix bug for write bench

Reviewed-by: Sage Weil <sage@redhat.com>
 2016-03-01 09:25:51 -05:00
Sage Weil
48e2a65b22 Merge pull request #7729 from yuyuyu101/wip-spdk2 
osd: bluestore: Revert NVMEDevice task cstor and refresh interface changes
 2016-03-01 09:18:15 -05:00
Nathan Cutler
b87f4efb56 packaging: lsb_release build and runtime dependency 
The lsb_release executable is being run in multiple places, not least in
src/common/util.cc, which calls it via shell in the collect_sys_info() code
path.

This patch addresses this issue on SUSE- and Debian-derivatives, as well
as reinstating the dependency for RHEL/Fedora after it was dropped in
1560057226.

http://tracker.ceph.com/issues/14906 Fixes: #14906

Signed-off-by: Nathan Cutler <ncutler@suse.com>
 2016-03-01 15:14:46 +01:00
Sage Weil
aa2b891140 buffer: use alignof for raw_combined allocation arithmetic 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 09:09:23 -05:00
Sage Weil
ef80690622 buffer: clean up raw_combined construction 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 09:09:22 -05:00
Sage Weil
b6ed4d3ebf buffer: size append_buffer so that it fits into page-multiple allocations 
We drop some unittest assertions about alloc buffer size.  Sorry!

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 09:09:22 -05:00
Sage Weil
f2c0d5cd10 rbd-replay: s/CEPH_BUFFER_APPEND_SIZE/CEPH_PAGE_SIZE/ 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 09:09:22 -05:00
Sage Weil
ce3e5a3347 buffer: alloc right-sized buffer from read_fd 
This may as well fit the input; this doesn't relate to the
append buffer.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 08:47:29 -05:00
Sage Weil
73dcd26fc8 buffer: use raw_combined for certain allocations 
If the alignment is on a page boundary, or the allocation is big,
a separate buffer::raw goes faster.  The rest of the time,
a raw_combined does.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 08:47:29 -05:00
Sage Weil
6be3b99d5d buffer: align unspecified allocations to a word 
This will let us put policy create_aligned.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 08:47:28 -05:00
Sage Weil
724a493adc buffer: combine data and buffer::raw into single allocation 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 08:47:28 -05:00
Sage Weil
04482ae902 buffer: add front(), back(), get_num_buffers() methods 
These eliminate most callers of buffers(), which exposes the
internal list<ptr>.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 08:47:28 -05:00
Sage Weil
69bcbe104d unittest_bufferlist: benchmark some allocations 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 08:47:28 -05:00
Sage Weil
08c0d98bc1 unittest_bufferlist: fix append_bench 
- fix source
- include larger sizes

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 08:47:28 -05:00
Sage Weil
7723d294a5 unittest_bufferlist: fix ptr move test 
Do not assume there is a trailing null the terminate the string.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-01 08:47:28 -05:00
Sage Weil
149f01f3d8 Merge pull request #7739 from xiexingguo/xxg-wip-14387 
osd: fix race condition for heartbeat_need_update

Reviewed-by: Sage Weil <sage@redhat.com>
 2016-03-01 08:46:08 -05:00
Sage Weil
ef5973324b Merge pull request #7703 from xiexingguo/xxg-wip-statfs 
osd: filestore: fast abort if statfs encounters ENOENT

Reviewed-by: Sage Weil <sage@redhat.com>
 2016-03-01 08:45:39 -05:00
Sage Weil
57b838f4b8 Merge pull request #7766 from adamemerson/wip-portable-mutex 
common: Do not use non-portable constants in mutex_debug

Reviewed-by: Willem Jan Withagen <wjw@digiware.nl>
 2016-03-01 08:45:00 -05:00
Sage Weil
5f429178b5 Merge pull request #7799 from xiexingguo/xxg-wip-fixnvme 
osd: bluestore: NVMEDevice: fix error handling

Reviewed-by: Sage Weil <sage@redhat.com>
 2016-03-01 08:44:21 -05:00
Sage Weil
46bd7bbb8b Merge pull request #7804 from jjhuo/fixes 
osd: kstore: fix a race condition in _txc_finish()

Reviewed-by: Sage Weil <sage@redhat.com>
 2016-03-01 08:43:49 -05:00
Sage Weil
0edaae6305 Merge pull request #7816 from marcan/build-fixes 
configure.ac: boost_iostreams is required, not optional

Reviewed-by: Kefu Chai <kchai@redhat.com>
 2016-03-01 08:43:25 -05:00