Sage Weil [Tue, 22 Oct 2013 19:54:09 +0000 (12:54 -0700)]
ceph: catch exceptions thrown during the rados handle init
In my case, making ceph.conf unreadable triggers an exception here:
Traceback (most recent call last):
File "./ceph", line 802, in <module>
sys.exit(main())
File "./ceph", line 575, in main
conf_defaults=conf_defaults, conffile=conffile)
File "/home/sage/src/ceph/src/pybind/rados.py", line 221, in __init__
self.conf_read_file(conffile)
File "/home/sage/src/ceph/src/pybind/rados.py", line 272, in conf_read_file
raise make_ex(ret, "error calling conf_read_file")
rados.Error: error calling conf_read_file: errno EACCES
Yehuda Sadeh [Mon, 21 Oct 2013 21:17:12 +0000 (14:17 -0700)]
rgw: turn swift COPY into PUT
Fixes: #6606
The swift COPY operation is unique in a sense that it's a write
operation that has its destination not set by the URI target, but by a
different HTTP header. This is problematic as there are some hidden
assumptions in the code that the specified bucket/object in the URI is
the operation target. E.g., certain initialization functions, quota,
etc. Instead of creating a specialized code everywhere for this case
just turn it into a regular copy operation, that is, a PUT with
a specified copy source.
Greg Farnum [Fri, 18 Oct 2013 23:34:11 +0000 (16:34 -0700)]
ReplicatedPG: copy: conditionally requeue copy ops when cancelled
We may need to requeue copy ops which are cancelled as part of an acting
set change but don't change the primary. To support this, add a
"requeue" flag to cancel_copy_ops() and copy_ops(), as well as to
CopyResults. The CopyCallback is then responsible for requeuing (the
higher layers can't do so as they can't know which request actually
triggered the copy).
Sage Weil [Thu, 17 Oct 2013 19:06:26 +0000 (12:06 -0700)]
Makefile: fix /sbin vs /usr/sbin behavior
Instead of telling configure to put things in /sbin, explicitly put the
two important items (mkcephfs and mount.fuse.ceph) in /sbin via an
automake rule. This unbreaks FreeBSD 9.1 and probably others.
Based on patches originally from Alan Somers <asomers@gmail.com>, modified
for the current Makefile structure and applied to the specfile too.
Fixes: #6456 Signed-off-by: Sage Weil <sage@inktank.com> Tested-by: Alan Somers <asomers@gmail.com>
Gregory Farnum [Wed, 16 Oct 2013 18:13:35 +0000 (11:13 -0700)]
Merge pull request #709 from ceph/wip-filerecover
This patch prevents us from inadvertently reducing sparse file sizes during recovery.
We also reduce some code duplication by using eval() directly in do_file_recover()
instead of reproducing the parts we care about.
Sage Weil [Fri, 27 Sep 2013 22:29:13 +0000 (15:29 -0700)]
common/buffer: add crc caching performance test
On my old box:
- matching cached values is a big win (free), obviously
- the adjustment is the same speed as redoing the calculation. this
is probably because the data is already in L1/L2 cache; we still
save memory bandwidth.
Sage Weil [Wed, 16 Oct 2013 00:55:32 +0000 (17:55 -0700)]
cls_rbd: do not make noise in osd log on rbd removal
ubuntu@burnupi06:~$ tail -f /var/log/ceph/ceph-osd.1.log
2013-02-07 17:00:30.565749 7fdb09e6b700 0 <cls> cls/rbd/cls_rbd.cc:1615: error reading id for name 'sds': -2
2013-02-07 17:00:30.566301 7fdb0a66c700 0 <cls> cls/rbd/cls_rbd.cc:1521: error reading name to id mapping: -2
2013-02-07 17:03:54.085700 7fdb0a66c700 0 <cls> cls/rbd/cls_rbd.cc:1615: error reading id for name 'sdfsd': -2
2013-02-07 17:03:54.086143 7fdb09e6b700 0 <cls> cls/rbd/cls_rbd.cc:1521: error reading name to id mapping: -2
Fixes: #4047 Signed-off-by: Sage Weil <sage@inktank.com>
Sage Weil [Tue, 15 Oct 2013 22:50:16 +0000 (15:50 -0700)]
ceph_test_rados: fix snap remove vs rollback fix
In commit 55d279b98553ba4542219b126fc7159b20b18b1f we tried to fix a race
between rollback and snap removal, but got the logic wrong: we need to
prevent *snap removal* on in-use snaps, not prevent multiple rollbacks on
the same snap.
Fixes: #6254 (again) Signed-off-by: Sage Weil <sage@inktank.com> Reviewed-by: Samuel Just <sam.just@inktank.com>
Alan Somers [Tue, 15 Oct 2013 20:26:13 +0000 (13:26 -0700)]
os/FileStore: fix fiemap double-free(s)
If GenericFileStoreBackend::do_fiemap returns anything other than 0, then
fiemap will not be allocated. However,
GenericFileStoreBackend::detect_features will free fiemap regardless,
triggering an assertion in tcmalloc. The attached patch will fix the bug
by only freeing fiemap when necessary.
In my case, do_fiemap returns -ENOSYS because fiemap is not implemented
on FreeBSD. However, this bug could also happen on Linux when do_fiemap
returns -ENOMEM.
Fixes: #6504 Signed-off-by: Alan Somers <asomers@gmail.com> Reviewed-by: Sage Weil <sage@inktank.com>
Alan Somers [Tue, 15 Oct 2013 20:06:06 +0000 (13:06 -0700)]
ceph-dencoder: select_generated() should properly validate its input
If m_list.size() == 0, then calling select_generated(0) will result in
uninitialized data being assigned to m_object, which will cause a segfault
down the road. This patch fixes that.
To Reproduce:
$ ceph-dencoder type MWatchNotify select_test 0 encode decode
Segmentation fault (core dumped)
After the patch:
$ ./ceph-dencoder type MWatchNotify select_test 0 encode decode
error: invalid id for generated object
$ echo $?
1
Fixes: #6510 Signed-off-by: Alan Somers <asomers@gmail.com>
Sage Weil [Tue, 15 Oct 2013 18:22:16 +0000 (11:22 -0700)]
mon/PGMonitor: set floor below which we do not warn about objects/pg
If a cluster has very few objects, do not generate warnings when the
objects/pg for a pool diverges from the cluster average. This avoids
spurious errors when you have a relatively empty cluster and a lone pool
with a modest number of objects is too far off the (mostly meaningless)
cluster-wide average.
Also include a per-pool min so we ignore mostly-empty pools.
Fixes: #6521 Signed-off-by: Sage Weil <sage@inktank.com>
projects / ceph.git / log