Commit Graph

34832 Commits

Author SHA1 Message Date
Samuel Just
de0c7202a0 Merge pull request #2226 from athanatos/wip-8396
Wip 8396

Reviewed-by: Sage Weil <sage@redhat.com>
2014-08-07 15:56:18 -07:00
Samuel Just
2120f4bb6c OSD: move waiting_for_pg into the session structures
Each message belongs to a session.  Further, no ordering is implied
between messages which arrived on different sessions.  Breaking the
global waiting_for_pg structure into a per-session structure lets
us avoid the problem of taking a write lock on a global structure
(pg_map_lock) in get_pg_or_queue_for_pg at the cost of some
complexity in updating each session's waiting_for_pg structure when
we receive a new map (due to pg splits) or when we locally create
a pg.

Signed-off-by: Samuel Just <sam.just@inktank.com>
2014-08-07 15:54:38 -07:00
Sage Weil
f5d6818d37 Merge pull request #2227 from athanatos/wip-8625
osd_types: s/stashed/rollback_info_completed and set on create

Reviewed-by: Sage Weil <sage@redhat.com>
2014-08-07 15:42:27 -07:00
John Wilkins
71cb4d3c0d doc: Removed Debian reference from Upstart.
Signed-off-by: John Wilkins <john.wilkins@inktank.com>
2014-08-07 15:30:43 -07:00
John Wilkins
8357cae027 doc: Incorporated user feedback to clarify upgrade doc.
Signed-off-by: John Wilkins <john.wilkins@inktank.com>
2014-08-07 15:29:34 -07:00
John Wilkins
b6bf33cd33 doc: Added configuration discussion at end of gateway install with links to config.
Signed-off-by: John Wilkins <john.wilkins@inktank.com>
2014-08-07 14:15:45 -07:00
Samuel Just
11cefcab23 OSD::shutdown: actually drop sessions waiting on map
There might be messages for which we still don't have the
map.  Dispatching waiting won't actually help.

Signed-off-by: Samuel Just <sam.just@inktank.com>
2014-08-07 11:48:54 -07:00
Samuel Just
6aba942170 OSD: clear_session_waiting_on_map in ms_handle_reset
Signed-off-by: Samuel Just <sam.just@inktank.com>
2014-08-07 11:48:54 -07:00
Samuel Just
11d6e5ad84 OSD: rename session_waiting_for_map_lock to session_waiting_lock
This lock will also protect the waiting_for_pg structures in each
session.

Signed-off-by: Samuel Just <sam.just@inktank.com>
2014-08-07 11:48:54 -07:00
Samuel Just
20fd714a08 OSD: wake_pg_waiters outside of the pgmap write_lock, pg_lock
Signed-off-by: Samuel Just <sam.just@inktank.com>
2014-08-07 11:48:54 -07:00
Samuel Just
6e7b86a24f OSD: fix wake_pg_waiters revert error in _open_lock_pg
231fe1b685 reintroduced erroneously
this call to wake_pg_waiters.  All _create_lock_pg callers handle
calling wake_pg_waiters after the pg lock has been dropped.

Fixes: #8691
Signed-off-by: Samuel Just <sam.just@inktank.com>
2014-08-07 11:48:54 -07:00
Samuel Just
4260767973 osd_types: s/stashed/rollback_info_completed and set on create
Originally, this flag indicated that the object had already been stashed and
that therefore recording subsequent changes is unecessary.  We want to set it
on create() as well since operations like [create, writefull] should not need
to stash the object.

Fixes: #8625
Signed-off-by: Samuel Just <sam.just@inktank.com>
2014-08-07 11:47:52 -07:00
Sage Weil
f95e82b134 Merge remote-tracking branch 'gh/wip-filestore-bigxattr' 2014-08-07 11:21:06 -07:00
Wido den Hollander
a5ecf1515e powerdns: Update README with better markdown 2014-08-07 15:12:19 +02:00
Sage Weil
9bbd11497b Merge pull request #2220 from somnathr/wip-lock-leak-fix
RadosClient: Fixing potential lock leaks.

Backport: firefly
Reviewed-by: Sage Weil <sage@redhat.com>
2014-08-06 18:42:18 -07:00
Pavan Rallabhandi
f1aad8bcfc RadosClient: Fixing potential lock leaks.
In lookup_pool and pool_delete, a lock is taken
before invoking wait_for_osdmap, but is not
released for the failure case of the call. Fixing the same.

Fixes: #9022
Signed-off-by: Pavan Rallabhandi <pavan.rallabhandi@sandisk.com>
2014-08-06 18:30:01 -07:00
Sage Weil
26750fcfe8 os/FileStore: force any new xattr into omap on E2BIG
If we have a huge xattr (or many little ones), the _fgetattrs() for the
inline_set will fail with E2BIG.  The conditions later where we decide
whether to clean up the old xattr will then also fail.  We *will* put
the xattr in omap, but the non-omap version isn't cleaned up.

Fix this by setting a flag if we get E2BIG that the inline_set is known
to be incomplete.  In that case, take the conservative step of assuming
the xattr might be present and chain_fremovexattr().  Ignore any error
because it might not be there.

This is clearly harmless in the general case because it won't be there.
If it is, we will hopefully remove enough xattrs that the E2BIG
condition will go away (usually by removing some really big chained
xattr).

See original bug #7779.  With this in place, we can repair objects in
the broken state if we know the rados attr(s) that are responsible.
Usually that is user.rgw.manifset, and a rados get + set of the attr
will repair things.

Reviewed-by: Yehuda Sadeh <yehuda@inktank.com>
Reviewed-by: Samuel Just <sam.just@inktank.com>
Signed-off-by: Sage Weil <sage@redhat.com>
2014-08-06 17:28:45 -07:00
Yehuda Sadeh
d050cf6c82 Merge pull request #2218 from ceph/wip-rados-xattr
rados: fix get/setxattr commands up

Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>
2014-08-06 17:11:44 -07:00
Sage Weil
cc3112efd6 rados: use STD{IN,OUT}_FILENO for magic values
Signed-off-by: Sage Weil <sage@redhat.com>
2014-08-06 17:06:26 -07:00
Sage Weil
e3819b6216 qa/workunits/rados/test_rados_tool: add a few xattr tests
Signed-off-by: Sage Weil <sage@redhat.com>
2014-08-06 17:06:26 -07:00
Sage Weil
645c28a108 rados: optionally read setxattr value from stdin
Signed-off-by: Sage Weil <sage@redhat.com>
2014-08-06 15:31:19 -07:00
Sage Weil
59a715a39c rados: don't add \n to getxattr
Signed-off-by: Sage Weil <sage@redhat.com>
2014-08-06 15:31:15 -07:00
Sage Weil
ea731ae142 qa/workunits/cephtool/test.sh: fix 'ceph df ...' tests
Broken by ee2dbdb0f5 and friends.

Signed-off-by: Sage Weil <sage@redhat.com>
2014-08-06 13:16:49 -07:00
Sage Weil
0bd7c34140 PendingReleaseNotes: make note about rbd cache default change
Signed-off-by: Sage Weil <sage@redhat.com>
2014-08-06 11:37:22 -07:00
Josh Durgin
8d8b8b1128 Merge pull request #2123 from ceph/wip-rbd-flush
librbd: enable rbd cache by default; writethrough until flush

Reviewed-by: Josh Durgin <josh.durgin@inktank.com>
2014-08-06 11:28:56 -07:00
Sage Weil
7d820123d0 Merge pull request #2205 from ceph/wip-librbd-snap-meta
librbd: fix crash with a chain of flattened images

Reviewed-by: Sage Weil <sage@redhat.com>
2014-08-06 11:08:43 -07:00
Sage Weil
c0ef8b8746 rocksdb: fix i386 build
Signed-off-by: Sage Weil <sage@redhat.com>
2014-08-06 10:51:33 -07:00
Sage Weil
0e9f3759cc Merge remote-tracking branch 'gh/wip-test-ceph-disk' 2014-08-06 10:41:36 -07:00
Sage Weil
6168c3c988 Merge pull request #2215 from ceph/wip-kb
mon: clean up _kb fields in json and perf counter output

Reviewed-by: John Spray <john.spray@redhat.com>
2014-08-06 10:33:36 -07:00
Sage Weil
ee2dbdb0f5 mon/PGMonitor: remove {rd,wr}_kb from pool stat dumps
These fields are replaced with corresponding *_bytes fields.

Signed-off-by: Sage Weil <sage@redhat.com>
2014-08-06 10:33:04 -07:00
Sage Weil
4c8e3224ea mon: remove *_kb perf counters
This is an incompatible change.

Signed-off-by: Sage Weil <sage@redhat.com>
2014-08-06 10:33:02 -07:00
Sage Weil
3179bb107a mon/PGMonitor: add _bytes perf counters
Leave the _kb ones in place for now.

Backport: firefly
Signed-off-by: Sage Weil <sage@redhat.com>
2014-08-06 09:18:27 -07:00
Sage Weil
adb2791adf mon/PGMonitor: add _bytes fields for all usage dumps
Leave the _kb ones in place for now.

Backport: firefly
Signed-off-by: Sage Weil <sage@redhat.com>
2014-08-06 09:14:40 -07:00
Loic Dachary
b416446cb9 Merge pull request #2165 from dachary/wip-mailmap
mailmap updates
2014-08-06 17:23:17 +02:00
Sage Weil
895318c0cd README.md: word wrap
Signed-off-by: Sage Weil <sage@redhat.com>
2014-08-06 08:16:21 -07:00
Sage Weil
500b95e568 README: symlink from README.md
It looks better as markdown than rendered as text via the markdown tool,
so just symlink it.

Signed-off-by: Sage Weil <sage@redhat.com>
2014-08-06 08:15:35 -07:00
Sage Weil
25b9761e24 Merge pull request #2202 from xinxinsh/enable-rocksdb-log-level
Enable rocksdb log level

Reviewed-by: Sage Weil <sage@redhat.com>
2014-08-06 07:46:04 -07:00
Sage Weil
859944d3c9 test/osd/osd-test-helpers: mkdir -p for ceph-disk
Signed-off-by: Sage Weil <sage@redhat.com>
2014-08-05 16:48:30 -07:00
Sage Weil
818f5b7038 Merge pull request #2208 from lpabon/osd_dev_doc
Developer quick start guide

Reviewed-by: Sage Weil <sage@redhat.com>
2014-08-05 15:25:40 -07:00
Sage Weil
c2f58e6694 test/ceph-disk.sh: mkdir -p
Signed-off-by: Sage Weil <sage@redhat.com>
2014-08-05 15:11:18 -07:00
Samuel Just
3f9f135f37 Merge remote-tracking branch 'upstream/next' into wip-sam-testing
Conflicts:
	src/osd/OSD.cc
2014-08-05 14:06:23 -07:00
Luis Pabón
a923e2c9eb Renamed README to README.md to render in markdown
Signed-off-by: Luis Pabón <lpabon@redhat.com>
2014-08-05 14:51:16 -04:00
Luis Pabón
52cf693e8d Developer quick start guide
Signed-off-by: Luis Pabón <lpabon@redhat.com>
2014-08-05 14:48:47 -04:00
xinxin shu
c09036aca2 enable info_log_level config option for rocksdb
Signed-off-by: xinxin shu <xinxin.shu@intel.com>
2014-08-06 02:11:38 +08:00
John Wilkins
e6935dd461 Merge pull request #2206 from ceph/wip-8875
doc: be a bit more explicit about 'ceph-deploy new' in quickstart

Reviewed-by: John Wilkins <john.wilkins@inktank.com>
2014-08-05 10:21:58 -07:00
Alfredo Deza
66de51d1a5 be a bit more explicit about 'ceph-deploy new' in quickstart
Signed-off-by: Alfredo Deza <alfredo.deza@inktank.com>
2014-08-05 12:52:00 -04:00
Sage Weil
3876da15b6 Merge branch 'master' of github.com:ceph/ceph 2014-08-05 09:15:34 -07:00
Sage Weil
bc1d2b8328 Merge remote-tracking branch 'gh/wip-8880'
Conflicts:
	src/osd/OSD.cc
2014-08-05 09:15:12 -07:00
Alfredo Deza
19b9ee947d Merge pull request #2204 from osynge/wip-dont-mkdir-by-mistake2
Do not make directories by mistake.

Reviewed-by: Alfredo Deza <adeza@redhat.com>
2014-08-05 11:50:57 -04:00
Sage Weil
b8c1cabd94 Merge pull request #1883 from ceph/wip-msgr
messenger refactoring for xio

Reviewed-by: Greg Farnum <greg@inktank.com>
2014-08-05 08:50:05 -07:00