Sage Weil [Mon, 11 Feb 2013 14:23:54 +0000 (06:23 -0800)]
osd: update snap collections for sub_op_modify log records conditionaly
The only remaining caller is sub_op_modify(). If we do have a non-empty
op transaction, we want to do this update, regardless of what we think
last_backfill is (our notion may be not completely in sync with the
primary). In particular, our last_backfill may be the same object but
a different snapid, but the primary disagrees and is pushing an op
transaction through.
Instead, update the collections if we have a non-empty transaction.
Signed-off-by: Sage Weil <sage@inktank.com> Reviewed-by: Samuel Just <sam.just@inktank.com>
Sage Weil [Mon, 11 Feb 2013 00:59:48 +0000 (16:59 -0800)]
osd: unconditionally encode snaps buffer
Previously we would only encode the updated snaps vector for CLONE ops.
This doesn't work for MODIFY ops generated by the snap trimmer, which
may also adjust the clone collections. It is also possible that other
operations may need to populate this field in the future (e.g.,
LOST_REVERT may, although it currently does not).
Fixes: #4071, and possibly #4051. Signed-off-by: Sage Weil <sage@inktank.com> Reviewed-by: Samuel Just <sam.just@inktank.com>
Joe Buck [Sun, 10 Feb 2013 02:48:57 +0000 (18:48 -0800)]
libcephfs: fix for #4068
If client->init() fails in mount, then client->shutdown()
should not be called. This patch uses a bool to ensure
that shutdown is only called if init() succeeds.
Signed-off-by: Joe Buck <jbbuck@gmail.com> Reviewed-by: Sam Lang <sam.lang@inktank.com> Reviewed-by: Sage Weil <sage@inktank.com>
ceph_authtool.cc:15:17: warning: using directive refers to
implicitly-defined namespace 'std'
using namespace std;
^ Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
librados/librados.cc:15:17: warning: using directive refers to
implicitly-defined namespace 'std'
using namespace std;
^
Include what we need and use the related classes.
Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
Danny Al-Gaaf [Thu, 7 Feb 2013 19:16:46 +0000 (20:16 +0100)]
src/osd/PG.h: use empty() instead of size()
Fix warning for usage of objects.size(). Use empty() since it
should be prefered as it has, following the standard, a constant
time complexity regardless of the containter type. The same is not
guaranteed for size().
warning from cppchecker was:
[osd/PG.h:599]: (performance) Possible inefficient checking for
'objects' emptiness.
Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
Danny Al-Gaaf [Thu, 7 Feb 2013 19:14:34 +0000 (20:14 +0100)]
src/osd/OSD.h: use empty() instead of size()
Fix warning for usage of *.size(). Use empty() since it should be
prefered as it has, following the standard, a constant time
complexity regardless of the containter type. The same is not
guaranteed for size().
warning from cppchecker was:
[osd/OSD.h:265]: (performance) Possible inefficient checking for
'last_scrub_pg' emptiness.
[osd/OSD.h:274]: (performance) Possible inefficient checking for
'last_scrub_pg' emptiness.
Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
Danny Al-Gaaf [Thu, 7 Feb 2013 19:04:39 +0000 (20:04 +0100)]
common/WorkQueue.h: use empty() instead of size()
Fix warning for usage of out->size() in _void_dequeue(). Use empty()
since it should be prefered as it has, following the standard, a
constant time complexity regardless of the containter type. The
same is not guaranteed for size().
warning from cppchecker was:
[common/WorkQueue.h:97]: (performance) Possible inefficient
checking for 'queue' emptiness.
Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
Danny Al-Gaaf [Thu, 7 Feb 2013 18:55:52 +0000 (19:55 +0100)]
common/AsyncReserver.h: use empty() instead of size()
Fix warning for usage of queue.size() in do_queues(). Use empty()
since it should be prefered as it has, following the standard, a
constant time complexity regardless of the containter type. The
same is not guaranteed for size().
warning from cppchecker was:
[common/AsyncReserver.h:40]: (performance) Possible inefficient
checking for 'queue' emptiness.
Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
Danny Al-Gaaf [Wed, 6 Feb 2013 15:03:09 +0000 (16:03 +0100)]
common/config.h: declaration of config_option as struct
Change declaration of config_option from 'class' to 'struct' since
it's defined as struct and used this way (access members). The declaration
as class doesn't change the behaviour.
Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
Danny Al-Gaaf [Sun, 10 Feb 2013 19:52:21 +0000 (20:52 +0100)]
.gitignore: cleanup and sort entries
Cleanup old entries and sort the existing entries. Remove no longer
existing entries/files. Move some entries from top level .gitignore
files down to the correct subdir .gitignore file.
Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
Danny Al-Gaaf [Sun, 10 Feb 2013 18:00:40 +0000 (19:00 +0100)]
ceph-test.install: updated filelist
Add some files which are part of the ceph-test package on RPM based
distributions already: ceph_test_mon_workloadgen, ceph_rgw_jsonparser,
ceph_test_objectcacher_stress.
Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
Danny Al-Gaaf [Sun, 10 Feb 2013 17:33:56 +0000 (18:33 +0100)]
ceph-test: give binaries from ceph-test package useful names
The files from the ceph-test subpackage are installed to /usr/bin,
give them more useful names to make sure that the user know they
belong to ceph. add a 'ceph_' prefix and change some test* binaries
to ceph_test_*.
Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de> Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
Loic Dachary [Sun, 10 Feb 2013 13:23:36 +0000 (14:23 +0100)]
buffer::ptr self assignment bug + patch
After
buffer::ptr a(1);
a = a;
a call to a.get_raw() will return a null pointer and there will be
no pointer referencing the original buffer::raw object although its
reference count is 1.
buffer::ptr& buffer::ptr::operator= (const ptr& p) is modified to use
a local buffer::raw pointer to fix the memory leak. a = a
is a noop instead of loosing the original raw buffer.
A set of unit tests is added src/test/bufferlist.cc to demonstrate
that the proposed change works as expected. It is checked with
valgrind that reports no memory leak. The same test can be run against
the original code to show that it leaks.
Sage Weil [Sat, 9 Feb 2013 05:36:13 +0000 (21:36 -0800)]
java: make CephMountTest use user.* xattr names
Changes to the xattr code in Ceph require
a few tweaks to existing test cases.
Specifically, there is now a ceph.file.layout
xattr by default and user defined xattrs
are prepended with "user."
Signed-off-by: Sage Weil <sage@inktank.com> Reviewed-by: Joe Buck <jbbuck@gmail.com> Reviewed-by: Noah Watkins <noahwatkins@gmail.com>
Yehuda Sadeh [Fri, 8 Feb 2013 21:14:49 +0000 (13:14 -0800)]
rgw: change json formatting for swift list container
Fixes: #4048
There is some difference in the way swift formats the
xml output and the json output for list container. In
xml the entity is named 'name' and in json it is named
'subdir'.
Sage Weil [Sat, 9 Feb 2013 08:05:33 +0000 (00:05 -0800)]
osd: fix load_pgs collection handling
On a _TEMP pg, is_pg() would succeed, which meant we weren't actually
hitting the cleanup checks. Instead, restructure this loop as positive
checks and handle each type of collection we understand.
This fixes _TEMP cleanup.
Signed-off-by: Sage Weil <sage@inktank.com> Reviewed-by: Samuel Just <sam.just@inktank.com>
Sage Weil [Sat, 9 Feb 2013 08:04:29 +0000 (00:04 -0800)]
osd: fix load_pgs handling of pg dirs without a head
If there is a pgid that passes coll_t::is_pg() but there is no head, we
will populate the pgs map but then fail later when we try to do
read_state. This is a side-effect of 55f8579.
Take explicit note of _head collections we see, and then warn when we
find stray snap collections.
Signed-off-by: Sage Weil <sage@inktank.com> Reviewed-by: Samuel Just <sam.just@inktank.com>
Sage Weil [Fri, 8 Feb 2013 06:06:14 +0000 (22:06 -0800)]
mon: handle -EAGAIN in completion contexts
We can get ECANCELED, EAGAIN, or success out of the completion contexts,
but in the EAGAIN case (meaning there was an election) we were sending
a success to the client. This resulted in client hangs and all-around
confusion when the monitor cluster was thrashing.
Backport: bobtail Signed-off-by: Sage Weil <sage@inktank.com> Reviewed-by: Joao Luis <joao.luis@inktank.com>
Yehuda Sadeh [Sat, 15 Dec 2012 01:29:37 +0000 (17:29 -0800)]
rgw: stream get_obj operation
Fixes: #2941
Instead of iterating through the parts one by one when reading
an object, we can now send multiple requests in parallel. Two new
configurables added to control the max request size, and the total
size of pending requests.
Samuel Just [Thu, 7 Feb 2013 19:53:28 +0000 (11:53 -0800)]
ReplicatedPG: check store for temp collection in have_temp_coll
We may not have "created" the temp collection since OSD restart
before removing the PG. have_temp_coll must also look at the
OSD store. Currently, the only user is pg removal, so the
extra work is acceptable.
Backport: bobtail Signed-off-by: Samuel Just <sam.just@inktank.com>
Yehuda Sadeh [Thu, 7 Feb 2013 00:43:48 +0000 (16:43 -0800)]
rgw: bucket recreation should not clobber bucket info
Fixes: #4039
User's list of buckets is getting modified even if bucket already
exists. This fix removes the newly created directory object, and
makes sure that user info's data points at the correct bucket.
Greg Farnum [Thu, 7 Feb 2013 18:25:33 +0000 (10:25 -0800)]
mds: rename mds_traceless_replies to mds_inject_traceless_reply_probability
Sage pointed out we should try for consistent naming on these debug
options, and this option is like our other inject_[x] stuff.
While we're here, make the config option comment a bit clearer.
Sage Weil [Thu, 7 Feb 2013 18:21:49 +0000 (10:21 -0800)]
osd: flush peering queue (consume maps) prior to boot
If the osd itself is behind on many maps during boot, it will get more and
(as part of that) flush the peering wq to ensure the pgs consume them.
However, it is possible for OSD to have latest/recnet maps, but pgs to be
behind, and to jump directly to boot and join. The OSD is then laggy and
unresponsive because the peering wq is way behind.
To avoid this, call consume_map() (kick the peering wq) at the end of
init and flush it to ensure we are *internally* all caught up before we
consider joining the cluster.
I'm pretty sure this is the root cause of #3905 and possibly #3995.
Signed-off-by: Sage Weil <sage@inktank.com> Reviewed-by: Josh Durgin <josh.durgin@inktank.com> Reviewed-by: Samuel Just <sam.just@inktank.com>
Greg Farnum [Thu, 7 Feb 2013 00:35:49 +0000 (16:35 -0800)]
client: trigger the completion in _flush when short-cutting
We missed a shortcut return from _flush() when doing e9a6694d0151b79c3a3b44cee5df8e3d4dcbfc2c, so _fsync() calls
were failing. To fix, if _flush discovers there's nothing to flush,
trigger the completion by calling onfinish->finish().
Fixes #4038
Signed-off-by: Greg Farnum <greg@inktank.com> Reviewed-by: Sage Weil <sage@inktank.com>
Greg Farnum [Wed, 6 Feb 2013 21:32:53 +0000 (13:32 -0800)]
mds: add "mds traceless replies" debug option
This option specifies (in the range 0-1) the percentage of modifying
operations that should be responded to without including a trace
for the dentries.
In order to implement it, we add a "modify" parameter to
set_trace_dist(), and the callers fill that with
mdr->req->may_write().
Signed-off-by: Greg Farnum <greg@inktank.com> Reviewed-by: Sam Lang <sam.lang@inktank.com>
Greg Farnum [Wed, 6 Feb 2013 18:22:06 +0000 (10:22 -0800)]
test: fix Throttle unit test.
A bunch of these are slightly racy so they're enclosed in loops. This
particular one, though, changes the Throttle state in ways that
invalidate the asserts. To fix, reset the state before commencing a
rerun.
Danny Al-Gaaf [Wed, 6 Feb 2013 11:17:02 +0000 (12:17 +0100)]
rbd-fuse: fix for loop in open_rbd_image()
Remove uninitialized usage of 'int i' as i++ from 'for' loop.
The variale 'i' is never used in this loop and initialized
before the next use with 0.
Related warning from clang++:
rbd_fuse/rbd-fuse.c:141:36: warning: variable 'i' is uninitialized
when used here [-Wuninitialized]
for (im = rbd_images; im != NULL; i++, im = im->next) {
Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
Fix "out-of-line declaration of a member must be a definition
[-Wout-of-line-declaration]". Remove ceph::crypto::shutdown() outside
the crypto related ifdef's. Without nss or cryptopp configure will
fail anyways.
Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>