RepoMirrors/ceph
1
0
Fork 0
mirror of https://github.com/ceph/ceph synced 2024-12-17 00:46:05 +00:00
Code Issues Packages Projects Releases Wiki Activity
10,043 Commits 211 Branches 394 Tags 2 GiB
07e1a2f6d3
Commit Graph

10043 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Sage Weil
d5c892766d sample.ceph.conf: add example on writeahead (vs parallel) journal 2010-04-21 14:24:26 -07:00
Sage Weil
f2160f1ab4 osd: create 'rbd' pool 2010-04-21 14:24:20 -07:00
Sage Weil
8478ebedde debian: require hdparm, recommend btrfs-tools 2010-04-21 10:58:42 -07:00
Sage Weil
d4bf8eefa9 debian: require hdparm, recommend btrfs-tools 2010-04-21 10:58:29 -07:00
Sage Weil
476606f483 mds: always use get_newest_seq (not snap) when doing cow 2010-04-20 16:39:13 -07:00
Sage Weil
327432aead mds: rename move_to_realm(), add debugging (cleanup) 2010-04-20 16:39:13 -07:00
Sage Weil
c719f4e442 mds: don't cow already cowed dentry 2010-04-20 16:39:13 -07:00
Sage Weil
93058af345 sample.ceph.conf: updated 2010-04-20 16:38:53 -07:00
Sage Weil
ba5a0d1ff6 fetch_config: enable fetching ceph.conf from a remote location 
fetch_config: always fetch if script exists and is executable

fetch_config: make /etc/ceph/sample.fetch_config +x

Makefile: fix whitespace
 2010-04-20 16:38:48 -07:00
Sage Weil
5a5cc18374 osd: don't capture SIGINT/SIGTERM; journal and/or btrfs snaps are sufficient. 
We used to do this to avoid corrupting the filestore, but since we can now
either roll forward with the writeahead journal OR roll back using btrfs
snaps, this is useless.  It wasn't a full solution anyway.
 2010-04-20 10:47:10 -07:00
Sage Weil
f7f6daf1d2 auth: less noisy debug 2010-04-20 10:39:23 -07:00
Sage Weil
34d308770f mds: drop mutation locks 
Saw this one, but fixed others that appear broken as well:

mds/MDCache.h: In function 'virtual Mutation::~Mutation()':
mds/MDCache.h:115: FAILED assert(pins.empty())
 1: (Mutation::~Mutation()+0x46) [0x791716]
 2: (MDCache::truncate_inode_logged(CInode*, Mutation*)+0x7b) [0x7bd6b7]
 3: (C_MDC_TruncateLogged::finish(int)+0x2c) [0x84100e]
 4: (finish_contexts(std::list<Context*, std::allocator<Context*> >&, int)+0x1af) [0x74a807]
 5: (Journaler::_finish_flush(int, long, utime_t, bool)+0x873) [0x92ea71]
 6: (Journaler::C_Flush::finish(int)+0x43) [0x936503]
 7: (Objecter::handle_osd_op_reply(MOSDOpReply*)+0xcf5) [0x8ff00f]
 8: (MDS::_dispatch(Message*)+0x1ee0) [0x726372]
 9: (MDS::ms_dispatch(Message*)+0x2f) [0x7274b9]
 10: (Messenger::ms_deliver_dispatch(Message*)+0x55) [0x719949]
 11: (SimpleMessenger::dispatch_entry()+0x4f4) [0x7074b2]
 12: (SimpleMessenger::DispatchThread::entry()+0x29) [0x7039d1]
 13: (Thread::_entry_func(void*)+0x20) [0x71412e]
 14: /lib/libpthread.so.0 [0x7f25474e073a]
 15: (clone()+0x6d) [0x7f254670a69d]
 2010-04-20 09:58:53 -07:00
Sage Weil
d8f1de5e87 osd: fix snapset loading from _snapdir 
Another masked variable!  In the same block as the last one!  Aie.
 2010-04-19 20:43:22 -07:00
Sage Weil
1915f2d590 mds: clone dentry for multiversion dir if linkage is changing... 
...even if the inode itself doesn't need to be cowed.  In particular, we
do the pre_cow_old_inode() thing, so it frequently doesn't need to be cowed,
but the dentry does when we are say unlinking a directory or some such.
 2010-04-19 20:29:14 -07:00
Sage Weil
07c7873dc9 mds: avoid null deref when hiding .ceph dir 2010-04-19 16:38:53 -07:00
Greg Farnum
4f402cb45e mds: rename send_message_client to send_message_client_counted 2010-04-19 14:32:36 -07:00
Greg Farnum
09366cbf21 mds: fix erroneous uses of send_message that need to be send_message_client 2010-04-19 14:22:50 -07:00
Greg Farnum
3484b84403 mds: Add new Connection-based send_message_client that uses inc_push_seq too 2010-04-19 14:22:23 -07:00
Jim Schutt
65ec0241ab autoconf: Fix detection of sync_file_range. 
Without this patch, on CentOS 5.4 ./configure reports that
sync_file_range is missing, but HAVE_SYNC_FILE_RANGE ends
up being defined in src/acconfig.h anyway.

Compile tested on CentOS 5.4 (which does not have sync_file_range(2)
in distro glibc) and Fedora 11 (which does).

Signed-off-by: Jim Schutt <jaschut@sandia.gov>
 2010-04-16 15:55:43 -07:00
Greg Farnum
a0e7d8dfba remove references to [group] and [mount] in sample.ceph.conf 2010-04-16 15:51:55 -07:00
Greg Farnum
d2af75de11 add executables to .gitignore 2010-04-16 15:18:26 -07:00
Greg Farnum
edf935d1ca mds: Fix bad pointer dereference brought about by switch to Message ref-counting 2010-04-16 14:55:07 -07:00
Sage Weil
d8be246ba3 mds: xlock versionlock on rename if witnesses 
This ensures that we don't pipeline dentry linkage updates when there
are witnesses.  That can cause problems because replicas don't see
projected dentry linkage info, and will get confused when they look at
the replica of the srcdn and it's, say, NULL and not srci.
 2010-04-16 13:52:47 -07:00
Greg Farnum
06e5528a1d mds: use Connection-based send_message in handle_discover 2010-04-16 12:30:44 -07:00
Greg Farnum
6a0d0d0612 msg: Remove "asker" data member from MDiscover. 
Fix uses.
 2010-04-16 12:28:56 -07:00
Greg Farnum
751a4056c7 mds: Fix some compile-time issues. 2010-04-16 12:28:14 -07:00
Greg Farnum
da13000d96 mds: Subsystems generally use the connection-based send_message when possible. 
This isn't a very deep change, though -- a lot of the remaining instances
could probably be fixed up just by changing passthrough variables and whatnot.
 2010-04-16 11:12:30 -07:00
Greg Farnum
05b0e3a5ea mds: create send_message(Message *m, Connection *c) passthrough 2010-04-16 09:36:30 -07:00
Yehuda Sadeh
b4d423151a rados: aio, reorder callback completion handler creation 2010-04-15 17:04:43 -07:00
Greg Farnum
bd34079e86 mds: set trivial uses of send_message to use the Connection version. 2010-04-15 16:34:17 -07:00
Greg Farnum
69682ad837 osd: send_message via Connection when possible 2010-04-15 15:58:18 -07:00
Greg Farnum
b38cebf074 osd: send reply via connection for MOSDSubOps 2010-04-15 15:58:18 -07:00
Greg Farnum
f6d2666492 osd: Send messages via Connection * when possible. 
Also, rename a message shadowing its trigger message so that we can.
 2010-04-15 15:58:18 -07:00
Greg Farnum
85b67cecdf msgr: Implement new send_message[_lazy](Message *m, Connection *con) functions. 
These can reduce the number of lookups and should be able to recover
from any pipes being destroyed while the Connection is still active.
 2010-04-15 15:58:13 -07:00
Sage Weil
47f5315fca filestore: be thorough about wiping out old osd data content during mkfs 
Delete any subvols or directories that we fine.
 2010-04-15 14:44:17 -07:00
Greg Farnum
057897e468 msgr: if a pipe** is passed in to submit_message, it's used. 
This is a double pointer because it's possible that a Pipe has
been destroyed since the send_message call, so we want to check
that the Pipe * remains set after we've taken the main lock.
 2010-04-15 13:43:16 -07:00
Greg Farnum
731414b2c0 msgr: give submit_message new interface. 2010-04-15 13:43:15 -07:00
Greg Farnum
6c9bee4910 msgr: Add a void *pipe to Connection. 
This is set on Connection creation and zeroed out in ~Pipe().
 2010-04-15 13:43:15 -07:00
Greg Farnum
a9800e7335 msgr: remove unnecessary clone variable from submit_message 2010-04-15 13:43:11 -07:00
Sage Weil
06d032e824 filestore: log snap_create error before asserting success 2010-04-15 10:12:40 -07:00
Sage Weil
2cc0ec2937 mon: do not build mon to mon authorizer if cephx is disabled 2010-04-15 10:09:42 -07:00
Greg Farnum
ca74b4bb3d atomic_t: Add a new version based on libatomic_ops. 
Hopefully this version will take less CPU since it isn't running Spinlocks, but
we'll need to test for that.
Includes appropriate Makefile and configure changes.
 2010-04-14 16:54:48 -07:00
Greg Farnum
d1d6da3b4b atomic_t: Remove (never used) unstable commonc++-based version 2010-04-14 16:54:48 -07:00
Greg Farnum
69b849c4bd buffer: Include Spinlock.h. 
This is inherited from one implementation of atomic_t but not necessarily all.
Including the file here is necessary in the case of new, faster atomic_t
implementations that don't use Spinlocks.
 2010-04-14 16:54:48 -07:00
Sage Weil
198d710681 osd: fix source type checks (osd, mon) 
The peer type is on the connection, no need to look at Session.  Also, we
don't currently have Sessions for outgoing connections.
 2010-04-14 15:18:52 -07:00
Sage Weil
afcee3b07f osd: don't ->put() a null session in _share_map_incoming 2010-04-14 15:08:47 -07:00
Sage Weil
ec8471d623 osd: queue pg update under pg sequencer 2010-04-14 14:47:44 -07:00
Sage Weil
69b108f8ce osd: queue pg creation under pg sequencer 
Avoids pg log, metadata writes from being reordered.  Had to refactor
try_create_pg to be less annoying.
 2010-04-14 14:47:05 -07:00
Sage Weil
cfabc0208b mds: not dout(0) request latency 2010-04-14 12:40:58 -07:00
Sage Weil
09916d3508 qa: set -e in workunits 2010-04-14 12:40:58 -07:00