RepoMirrors/ceph
1
0
Fork 0
mirror of https://github.com/ceph/ceph synced 2025-02-01 16:03:43 +00:00
Code Issues Packages Projects Releases Wiki Activity
54,137 Commits 215 Branches 394 Tags 1.4 GiB
f75221c9a9
Commit Graph

54137 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Kefu Chai
f67ebef597 cmake: compile and install ceph-bluefs-tool 
Signed-off-by: Kefu Chai <kchai@redhat.com>
 2016-06-01 23:49:45 +08:00
Kefu Chai
6df0bc8691 cmake: s/ceph_test_xattr_bench/ceph_xattr_bench/ 
to match with automake and packager

Signed-off-by: Kefu Chai <kchai@redhat.com>
 2016-06-01 23:49:45 +08:00
Kefu Chai
81cd483936 cmake: rename ceph-psim to ceph_psim 
to match with automake and packager

Signed-off-by: Kefu Chai <kchai@redhat.com>
 2016-06-01 23:49:45 +08:00
Kefu Chai
b6da5c5135 cmake: fix bash_completion install path 
Signed-off-by: Kefu Chai <kchai@redhat.com>
 2016-06-01 23:49:45 +08:00
Sage Weil
1c2c6cc1d8 os/bluestore: fsck: check for dup overlay keys 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:40:50 -04:00
Sage Weil
380392fcec os/bluestore: fsck: use common helper to verify blobs and refs 
The checks are the same (or should be--we had missed a few).

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:40:50 -04:00
Sage Weil
cb7f0f75d3 os/bluestore: add FLAG_NOCACHE flag; do not cache unbuffered writes 
Add a Buffer flag to mark that a buffer should not be cached once it is
stable.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:40:50 -04:00
Sage Weil
029ced9680 os/bluestore: do not use buffered bdev in write path 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:40:50 -04:00
Sage Weil
b6d7c97daf os/bluestore: do not rely on bdev buffered reads in read path 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:40:50 -04:00
Sage Weil
247bcdbbfd os/bluestore: no buffered write in wal path 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:40:50 -04:00
Sage Weil
7679407378 os/bluestore: populate buffer cache on read 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:40:50 -04:00
Sage Weil
74dd955728 os/bluestore: keep intrusive_list of WRITING buffers 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:40:50 -04:00
Sage Weil
6e5249a4be os/bluestore: simple per-collection lru for buffers 
Size these using a global config.  This is only a starting point--we'll
obviously have to rework this to share memory across collections.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:40:50 -04:00
Sage Weil
63279896a0 os/bluestore: use bufferptr for csum_data 
encode/decode of vector<char> is not optimized.  Bufferptr is a more
natural type here anyway.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:40:50 -04:00
Igor Fedotov
6cec4f388d ceph_test_objecstore: Adds a test case for compression stuff verification (incomplete) 
Signed-off-by: Igor Fedotov <ifedotov@mirantis.com>
 2016-06-01 11:40:49 -04:00
Igor Fedotov
c307c0ffdd os/bluestore: Fixes configuration observation. 
Signed-off-by: Igor Fedotov <ifedotov@mirantis.com>
 2016-06-01 11:40:49 -04:00
Igor Fedotov
8f8fd3a722 os/bluestore: Cosmetic fixes in bluestore logging 
Signed-off-by: Igor Fedotov <ifedotov@mirantis.com>
 2016-06-01 11:40:49 -04:00
Igor Fedotov
e2f6a66ded os/bluestore: Enables cow for cloning at bluestore for store test 
Signed-off-by: Igor Fedotov <ifedotov@mirantis.com>
 2016-06-01 11:40:49 -04:00
Igor Fedotov
c7ed3aa7a2 os/bluestore: Fixes Bnode serialization/deserialization and removes legacy Bnode::ref_map 
Signed-off-by: Igor Fedotov <ifedotov@mirantis.com>
 2016-06-01 11:40:49 -04:00
Igor Fedotov
b0981b3391 ceph_test_objectstore: extends SimpleObjectTest with the case where write happens for neighboring csum blocks to verify for potential alignment issue 
Signed-off-by: Igor Fedotov <ifedotov@mirantis.com>
 2016-06-01 11:40:49 -04:00
Igor Fedotov
5361cb887f os/bluestore: Removes legacy block_size retrieval 
Signed-off-by: Igor Fedotov <ifedotov@mirantis.com>
 2016-06-01 11:40:49 -04:00
Sage Weil
326bb0f865 os/bluestore: use WriteContext and do_alloc_write for _do_write_small 
Kill some mostly-duplicated code

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:40:49 -04:00
Sage Weil
eb977e6ba3 os/bluestore: consolidate WriteContext items into a write_item 
Also include b_off in there.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:40:49 -04:00
Sage Weil
3d49c2eb57 os/bluestore: avoid unnecessary write_onode calls 
_wctx_finish callers always write the onode; we only need to worry about
our changes to the bnode.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:40:49 -04:00
Sage Weil
b0cabb78db os/bluestore: drop unused _pad_* methods 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:40:49 -04:00
Sage Weil
24578bc8f8 os/bluestore: drop unused _pad_zeros args 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:40:49 -04:00
Sage Weil
a9a5e63d99 os/bluestore: fix offset skew check 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:40:48 -04:00
Sage Weil
ea1a787c65 os/bluestore: ~0x -> ~ 
e.g., 0x432da000~1000 instead of 0x432da000~0x1000

I think it's sufficiently clear that the value after ~ should have the same
base as the first bit, and it's easier to read.  And less text.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:40:48 -04:00
Igor Fedotov
183db05a35 compressor: Extends decompressor interface to be able to provide compressed data length. 
Signed-off-by: Igor Fedotov <ifedotov@mirantis.com>
 2016-06-01 11:40:48 -04:00
Sage Weil
fe6aaca1db os/bluestore: compress on write 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:40:48 -04:00
Sage Weil
502473a95c os/bluestore: do not partially deallocate compressed blobs 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:40:48 -04:00
Sage Weil
931264ecab os/bluestore: _do_write_big: limit size of blobs based on compression mode 
We may want to compress in smaller chunks based on hints/policy.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:40:48 -04:00
Sage Weil
53b73328dd os/bluestore: track new compression config options 
Class-wide Compressor, compression mode, and options.  For now these are
global, although later we'll do them per-Collection so they can be pool-
specific.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:40:48 -04:00
Sage Weil
a7c9c84eac os/bluestore/bluestore_types: add length to the compression_header_t 
Snappy fails to decompress if there are extra zeros in the input buffer.
So, store the length explicitly in the header to avoid feeding them into
the decompressor.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:40:48 -04:00
Sage Weil
c4f1facaa5 os/bluestore: fix BufferSpace::read() 
- we weren't reading from 'clean' buffers
- restructured loop a bit chasing another bug (but it ended up being
  in the caller)

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:40:48 -04:00
Sage Weil
6977d28863 librados: add COMPRESSIBLE and INCOMPRESSIBLE alloc hints 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:40:48 -04:00
Sage Weil
9aec0a7a5a compressor: add a get_type() method to Compressor interface 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:40:47 -04:00
Sage Weil
53b699edda os/bluestore: fix _do_read cached vs read result assembly 
We weren't handling the case of

 read block 0~300
 cache bloc 100~100

where the result is read(head) + cached + read(tail). Restructure the
loop to handle this.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:40:47 -04:00
Sage Weil
d315a21be9 os/bluestore: fix _do_read read out of buffer cache 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:40:47 -04:00
Sage Weil
32c6ba129d os/bluestore: fix up _set_csum helper 
- make it thread-safe
- call during mount

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:40:47 -04:00
Igor Fedotov
f4c8d845e6 os/store_test: Fixes dump_mismatch_bl to avoid assert on lengths mismatch. Starts using it for BufferCacheTest 
Signed-off-by: Igor Fedotov <ifedotov@mirantis.com>
 2016-06-01 11:40:47 -04:00
Sage Weil
fb45f389de os/bluestore: use bdev_block_size instead of min_alloc_size for allocators 
min_alloc_size is more dynamic; we just need the block size unit here.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:40:47 -04:00
Ramesh Chander
8185f2d356 os/bluestore: min_alloc_size options for different media types 
Signed-off-by: Ramesh Chander <Ramesh.Chander@sandisk.com>
 2016-06-01 11:40:47 -04:00
Igor Fedotov
6148e1e74a os/bluestore: Fixes duplicate blob move when cloning 
Signed-off-by: Igor Fedotov <ifedotov@mirantis.com>
 2016-06-01 11:38:54 -04:00
Sage Weil
8b417f346a os/bluestore: avoid passing overlapping allocated/released sets to fm 
BitmapFreelistManager doesn't like overlapping allocated+released sets
when the debug option is enabled, because it does a read to verify the
op is valid and that may not have been applied to the kv store yet.

This makes bluestore ObjectStore/StoreTest.SimpleCloneTest/2 pass with
bluestore_clone_cow = false and bluestore_freelist_type = bitmap.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:38:54 -04:00
Sage Weil
7c04c21574 os/bluestore/BitmapFreelistManager: drop newline on hex dumps 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:38:54 -04:00
Sage Weil
46522cf0d2 buffer: add no-newline hexdump option 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:38:53 -04:00
Sage Weil
7f6174e9d6 os/bluestore/BitmapFreelistManager: use hex 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:38:53 -04:00
Sage Weil
c97578070e os/bluestore: drop warning 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:38:53 -04:00
Sage Weil
0b80659a0f ceph_test_objectstore: fix BufferCacheReadTest 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-06-01 11:38:53 -04:00