ceph_test_rados_api_c_read_operations: do not assert per-op rval is correct

This is not included in the pg log and may be zeroed if there is an op
replayed.  Disable the assertion until the underlying bug is fixed.  See
http://tracker.ceph.com/issues/19518

Signed-off-by: Sage Weil <sage@redhat.com>

Merge pull request #16171 from wjwithagen/wip-wjw-freebsd-nobtrfs

os/filestore: Exclude BTRFS on FreeBSD

Reviewed-by: Sage Weil <sage@redhat.com>

Merge pull request #15670 from rzarzynski/wip-rgw-postobj-get_params-error-handling

rgw: fix error handling in get_params() of RGWPostObj_ObjStore_S3.

Reviewed-by: Casey Bodley <cbodley@redhat.com>

Merge pull request #15700 from mikulely/olh-cleanup

rgw: cleanup rgw-admin duplicated judge during OLH GET/READLOG

Reviewed-by: Casey Bodley <cbodley@redhat.com>

Merge pull request #15777 from oritwas/wip-rgw-no-log-meta

rgw: log_meta only for more than one zone

Reviewed-by: Casey Bodley <cbodley@redhat.com>

Merge pull request #15827 from joscollin/wip-CID1412979-uninitialized-members

rgw: Uninitialized member in LCRule

Reviewed-by: Casey Bodley <cbodley@redhat.com>

Merge pull request #15938 from shashalu/fix-data-sync-run

rgw: fix radosgw-admin data sync run crash

Reviewed-by: Casey Bodley <cbodley@redhat.com>

Merge pull request #15949 from joscollin/wip-cleanup-rgw-rest-warning

rgw: silence warning from -Wmaybe-uninitialized

Reviewed-by: Casey Bodley <cbodley@redhat.com>

Merge pull request #15960 from aclamk/fix-rgw-cryptopp-compilation

radosgw: fix compilation with cryptopp

Reviewed-by: Casey Bodley <cbodley@redhat.com>

Merge pull request #15965 from mdw-at-linuxbox/signature

rgw: fix s3 object uploads with chunked transfers and v4 signatures.

Reviewed-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
Reviewed-by: Matt Benjamin <mbenjami@redhat.com>

Merge pull request #16156 from tchaikov/wip-20524

osd/filestore: Revert "os/filestore: move ondisk in front

Reviewed-by: Pan Liu <liupan1111@gmail.com>

filestore: Exclude BTRFS on FreeBSD

Signed-off-by: Willem Jan Withagen <wjw@digiware.nl>

Merge pull request #16155 from majianpeng/man-rados-add-offset

doc/rados.8: add offset option for put command

Reviewed-by: Kefu Chai <kchai@redhat.com>

doc/rados.8: add offset option for put command.

Signed-off-by: Jianpeng Ma <jianpeng.ma@intel.com>

Merge pull request #16032 from tanghaodong25/bluestore-aio-submit

os/bluestore/KernelDevice: batch aio submit

Reviewed-by: Sage Weil <sage@redhat.com>

Merge pull request #16086 from liewegas/wip-btrfs-experimental

os/filestore: require experimental flag for btrfs

Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Reviewed-by: xie xingguo <xie.xingguo@zte.com.cn>

Revert "os/filestore: move ondisk in front, so that return oncommit earlier, and improve write op performance."

This reverts commit 21f167cf39cc7c025004e9b455ce8ba1252863f0.

Fixes: http://tracker.ceph.com/issues/20524
Signed-off-by: Kefu Chai <kchai@redhat.com>

Merge pull request #16026 from linuxbox2/wip-tag-timeout-secs

rgw: cls: ceph::timespan tag_timeout wrong units

Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>

Merge pull request #15873 from ztczll/zl

osd: check queue_transaction return value

Reviewed-by: Sage Weil <sage@redhat.com>
Reviewed-by: Jos Collin <jcollin@redhat.com>

Merge pull request #16048 from Yan-waller/wip-waller-0630mon

mon/mgr: add detail error infomation

Reviewed-by: John Spray <john.spray@redhat.com>

Merge pull request #16122 from tchaikov/wip-qa-warning-msg

qa/tasks/ceph: should be "Waiting for all PGs", not "all osds"

Reviewed-by: xie xingguo <xie.xingguo@zte.com.cn>

Merge pull request #16117 from majianpeng/bluestore-move-object-exist-in-assign-nid

os/bluestore: move object exist in assign nid

Reviewed-by: xie xingguo <xie.xingguo@zte.com.cn>
Reviewed-by: Sage Weil <sage@redhat.com>

os/bluestore/BlueStore: move "onode::exists=true" in _assign_nid.

If object  has nid, it always set exists =true in get_onode.

Signed-off-by: Jianpeng Ma <jianpeng.ma@intel.com>

Merge pull request #16110 from javacruft/wrap-and-sort

debian: wrap-and-sort all files

Reviewed-by: Kefu Chai <kchai@redhat.com>

Merge pull request #15642 from wjwithagen/wip-wjw-freebsd-ceph-disk-zap

ceph_disk/main.py: Allow FreeBSD zap a OSD disk

Reviewed-by: Loic Dachary <loic@dachary.org>
Reviewed-by: Kefu Chai <kchai@redhat.com>

Merge pull request #16084 from mikulely/fix-data-pool-cli

rgw,test: fix rgw placement rule pool config option

Reviewed By: Jos Collin <jcollin@redhat.com>

Merge pull request #16024 from dillaman/wip-19035

rbd: do not attempt to load key if auth is disabled

Reviewed-by: Mykola Golub <mgolub@mirantis.com>

Merge pull request #15860 from dillaman/wip-20168

librbd: fail IO request when exclusive lock cannot be obtained

Reviewed-by: Mykola Golub <mgolub@mirantis.com>

Merge pull request #16118 from IlsooByun/fix-spdk-build-error

os/bluestore: fix build errors when spdk is on

Reviewed-by: Haomai Wang <haomai@xsky.com>

qa/tasks/ceph: should be "Waiting for all PGs", not "all osds"

amend the warning message

Signed-off-by: Kefu Chai <kchai@redhat.com>

Merge pull request #15265 from cooldavid/pr-osd-historic-op

osd: Preserve OSDOp information for historic ops

Reviewed-by: Sage Weil <sage@redhat.com>
Reviewed-by: Kefu Chai <kchai@redhat.com>

Merge pull request #15964 from tchaikov/wip-5249

mon/MonClient: respect the priority in SRV RR

Reviewed-by: Gregory Farnum <gfarnum@redhat.com>

Merge pull request #15780 from wjwithagen/wip-wjw-freebsd-use-ceph_to_hostos_errno

core: make the conversion from wire error to host OS work

Reviewed-by: Sage Weil <sage@redhat.com>
Reviewed-by: Kefu Chai <kchai@redhat.com>

core: make the conversion from wire error to host OS work

 - The key change is the type of rval,
   that will call the conversion when en/decoded
 - Remainder is fixes for the type change and promotions

Signed-off-by: Willem Jan Withagen <wjw@digiware.nl>

Merge pull request #15897 from yuyuyu101/wip-20390

msg/async: increase worker reference with local listen table enabled backend

Reviewed-by: Kefu Chai <kchai@redhat.com>

Merge pull request #15754 from tchaikov/wip-test-auth-caps

qa/suites: add test exercising workunits/mon/auth_caps.sh

Reviewed By: Brad Hubbard <bhubbard@redhat.com>
Reviewed-by: Joao Eduardo Luis <joao@suse.de>

Merge pull request #16011 from hsepeng/cache-osd-bugfix

osd/PrimaryLogPG solve cache tier osd high memory consumption

Reviewed-by: Greg Farnum <gfarnum@redhat.com>
Reviewed-by: Sage Weil <sage@redhat.com>

Merge pull request #13918 from liupan1111/wip-fix-performance-filestore

os/filestore: queue ondisk completion before apply work

Reviewed-by: Sage Weil <sage@redhat.com>

Merge pull request #16051 from liewegas/wip-20381

os/bluestore: fix deferred_aio deadlock

Reviewed-by: xie xingguo <xie.xingguo@zte.com.cn>

os/bluestore: fix build errors when spdk is on

- remove pthread library link from dpdk cmake
- fix the bugs of NVMEDevice

Signed-off-by: Ilsoo Byun <ilsoo.byun@sk.com>

Merge pull request #15958 from liewegas/wip-mgr-modules

mgr,mon: enable/disable mgr modules via 'ceph mgr module ...' commands

Reviewed-by: John Spray <john.spray@redhat.com>

debian: wrap-and-sort all files

'wrap-and-sort -a -t' debian folder to apply consistent
ordering/formatting to all files.

This makes synchronization of debian package changes
between Ceph packages and Ubuntu packages easier to
manage.

Signed-off-by: James Page <james.page@ubuntu.com>

Merge pull request #15305 from wjwithagen/wip-wjw-HAVE_BABELTRACE

cmake: Rewrite HAVE_BABELTRACE  option to WITH_

Reviewed-by: Kefu Chai <kchai@redhat.com>

Merge pull request #16094 from zphj1987/master

pybind/mgr/dashboard: fix get kernel_version error

Reviewed-by: Kefu Chai <kchai@redhat.com>

ceph-mgr:fix ceph-mgr web get kernel_version error

fix :
File "/usr/lib64/ceph/mgr/dashboard/module.py", line 523, in _clients
client['version'] = client['kernel_version']
----------

change client['kernel_version'] to
client['client_metadata']['kernel_version']

Signed-off-by: Peng Zhang zphj1987@gmail.com

Merge pull request #16093 from tchaikov/wip-20432

osd: clear_queued_recovery() in on_shutdown()

Reviewed-by: Josh Durgin <jdurgin@redhat.com>

Merge pull request #15823 from jdurgin/wip-omap-tp-heartbeat

osd/ReplicatedBackend: reset thread heartbeat after every omap entry …

Reviewed-by: Sage Weil <sage@redhat.com>
Reviewed-by: xie xingguo <xie.xingguo@zte.com.cn>

osd: clear_queued_recovery() in on_shutdown()

otherwise when we are kicking the PG waiting in OSD's awaiting_throttle
queue, the queue is still holding a strong reference of it. then we will
have warnings like "pgid 0.7 has ref count of 2" when shutting down OSD,
because pgmap is also holding a reference of it.

we could clear the "awaiting_throttle" directly in OSD::shutdown(), but
clear_queued_recovery() is more consistent this way, because it is PG
which registered itself to the queue in OSD, it'd be better if PG can
remove itself from the OSD when it is shutting down itself.

Fixes: http://tracker.ceph.com/issues/20432
Signed-off-by: Kefu Chai <kchai@redhat.com>

Merge pull request #16081 from tsg-/tsg-affiliation

mailmap, organizationmap: add affiliation for Tushar Gohad

Reviewed-by: Jos Collin <jcollin@redhat.com>

os/filestore: require experimental flag for btrfs

This isn't strictly accurate since we're really deprecating, but the
message is appropriately scary and reflects the risk of data loss.

Signed-off-by: Sage Weil <sage@redhat.com>

Merge remote-tracking branch 'upstream/pull/16080/head' into HEAD

* upstream/pull/16080/head:
  doc/mon: clean up `mds fail`/`mds deactivate` descriptions

Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>

doc: fix rgw placment rule pool config option

actually we're checking '--data-pool' and '--data-extra-pool'
during the argument parsing.

Signed-off-by: Jiaying Ren <jiaying.ren@umcloud.com>

Merge pull request #15833 from Linbing/wip-osd-delete-MPing

osd: osd does not using MPing Messages,do not include unused include

Reviewed-by: Kefu Chai <kchai@redhat.com>

doc/mon: clean up `mds fail`/`mds deactivate` descriptions

Fixes: http://tracker.ceph.com/issues/20424
Signed-off-by: John Spray <john.spray@redhat.com>

Merge pull request #16072 from tchaikov/wip-20453

debian: workaround the bug in dpkg-maintscript-helper

Reviewed-by: Sage Weil <sage@redhat.com>

Merge pull request #15835 from SUSE/wip-flatten-systemd-target-hierarchy-master

systemd: Add explicit Before=ceph.target

Reviewed-by: Nathan Cutler <ncutler@suse.com>
Reviewed-by: Boris Ranto <branto@redhat.com>

Merge pull request #16067 from YangFanlinux/affi

 .mailmap, .organizationmap: Update Fan Yang information and affiliation

Reviewed By: Jos Collin <jcollin@redhat.com>

Merge pull request #16055 from dmick/master

ceph.in: allow developer mode from outside build tree

Reviewed-by: Sage Weil <sage@redhat.com>
Reviewed-by: Kefu Chai <kchai@redhat.com>

debian: workaround the bug in dpkg-maintscript-helper

see https://bugs.launchpad.net/ubuntu/+source/snap-confine/+bug/1605052
and https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=582819

Fixes: http://tracker.ceph.com/issues/20453
Signed-off-by: Kefu Chai <kchai@redhat.com>

mailmap: Fan Yang information and affiliation

Signed-off-by: Fan Yang <yangf@neunn.com>

mailmap: add affiliation for Tushar Gohad

Signed-off-by: Tushar Gohad <tushar.gohad@intel.com>

osd/PrimaryLogPG solve cache tier osd high memory consumption

during cache tier dirty data flushing, cache tier osd memroy consumption
will get increasing due to the accumulative pg log which will not be trimmed

Fixes: http://tracker.ceph.com/issues/20464
Signed-off-by: Peng Xie <peng.hse@xtaotech.com>

CMakeLists.txt: Rewrite 2 remaning options to WITH_

All options to en/disable inclusion of libraries or other software
are of the format WITH_ so that the Cmake commaind like ahs all
WITH_* options. The WITH_=ON option will result in a HAVE_ setting
in CMAKE so that tests can use that variable.

Last "abusers" to actually follow this format.

 - HAVE_BABELTRACE

 - HAVE_ZFSLIB is fixed in #15907

Signed-off-by: Willem Jan Withagen <wjw@digiware.nl>

ceph_disk/main.py: Allow FreeBSD zap a OSD disk

 - refactor zap() into a Linux and FreeBSD version

Signed-off-by: Willem Jan Withagen <wjw@digiware.nl>

Merge pull request #16056 from xiexingguo/wip-kill-log-warn

common: fix log warnings

Reviewed-by: Sage Weil <sage@redhat.com>

os/bluestore: fix deferred_aio deadlock

We submit aios while holding deferred_lock, and we take deferred_lock in
the aio completion handler (of which there is only 1).  That means that
if the aio queue fills up, the submitter will block with the lock while
the finisher won't be able to take the lock, progress, and drain completed
aios, leading to a deadlock.

Fix this by submiting aios *without* deferred_lock held.  Instead, demote
to a deferred_submit_lock.  Further, in the finisher, submit aios
via a finisher (this only happens when in deferred_aggressive mode which
is rare and not performance-sensitive).

Signed-off-by: Sage Weil <sage@redhat.com>

Merge pull request #15986 from liewegas/wip-20435

mon/MgrStatMonitor: avoid dup health warnings during luminous upgrade

Reviewed-by: Gregory Farnum <gfarnum@redhat.com>

Merge pull request #14178 from songbaisen/m10

mon: show the leader info on mon stat command

Reviewed-by: Joao Eduardo Luis <joao@suse.de>

Merge pull request #15585 from dachary/wip-20229-ceph-disk-timeout

ceph-disk: set the default systemd unit timeout to 3h

Reviewed-by: Josh Durgin <jdurgin@redhat.com>

Merge pull request #15984 from idryomov/wip-chooseargs-encoding-fix

crush/CrushWrapper: chooseargs encoding fix

Reviewed-by: Sage Weil <sage@redhat.com>

Merge pull request #16023 from liewegas/wip-19994

systemd/ceph-mgr: remove automagic mgr creation hack

Reviewed-by: Ken Dreyer <kdreyer@redhat.com>
Reviewed-by: John Spray <john.spray@redhat.com>

Merge pull request #15956 from dillaman/wip-librbd-devstack

test: fix failing rbd devstack teuthology test

Reviewed-by: Mykola Golub <mgolub@mirantis.com>

Merge pull request #15923 from dillaman/wip-qa-rbd-memcheck-python

qa/suites/rbd: restrict python memcheck validation to CentOS

Reviewed-by: Mykola Golub <mgolub@mirantis.com>

common: fix log warnings

/home/jenkins-build/build/workspace/ceph-pull-requests/src/common/LogEntry.cc:54:24: warning: suggest parentheses around comparison in operand of ‘|’ [-Wparentheses]
   } else if (level_str == "warn" | level_str == "warning") {
                        ^
/home/jenkins-build/build/workspace/ceph-pull-requests/src/common/LogEntry.cc:56:24: warning: suggest parentheses around comparison in operand of ‘|’ [-Wparentheses]
   } else if (level_str == "error" | level_str == "err") {

Signed-off-by: xie xingguo <xie.xingguo@zte.com.cn>

ceph.in: allow developer mode from outside build tree

Use path of script to look for 'am I in a build tree' rather than
getcwd(), so that 'build tree' versions can be run from anywhere

Fixes: http://tracker.ceph.com/issues/20472
Signed-off-by: Dan Mick <dan.mick@redhat.com>

Merge pull request #15911 from xiexingguo/wip-bluestore-20170626

os/bluestore: narrow cache lock range; make sure min_alloc_size p2 aligned

Reviewed-by: Igor Fedotov <ifedotov@mirantis.com>

Merge pull request #16044 from liewegas/wip-i-cant-believe-its-not-btrfs

qa: stop testing btrfs

Reviewed-by: Josh Durgin <jdurgin@redhat.com>

Merge pull request #16046 from liewegas/wip-valgrind-on-centos

do all valgrind runs on centos

Reviewed-by: Josh Durgin <jdurgin@redhat.com>

doc: recommend against btrfs

Signed-off-by: Sage Weil <sage@redhat.com>

mon/mgr: add detail error infomation

Signed-off-by: Yan Jun <yan.jun8@zte.com.cn>

Merge pull request #15963 from Songweibin/wip-release-notes

doc/release-notes: add Images creation timestamp note

Reviewed-by: xie xingguo <xie.xingguo@zte.com.cn>

Merge pull request #15924 from jcsp/wip-log-last-filters

mon: Filter `log last` output by severity and channel

Reviewed-by: Sage Weil <sage@redhat.com>
Reviewed-by: Kefu Chai <kchai@redhat.com>

Merge pull request #15946 from Yan-waller/wip-waller-dumpwatcher0627

osd: dump the field name of object watchers  and cleanups

Reviewed-by: xie xingguo <xie.xingguo@zte.com.cn>

Merge pull request #15959 from smithfarm/wip-rgw-singleton-cleanup

tests: rgw/singleton: drop duplicate filestore-xfs.yaml

Reviewed-by: Casey Bodley <cbodley@redhat.com>

Merge pull request #15961 from majianpeng/filestore-remove-redundant-judgement

os/filestore/FileStore.cc: remove a redundant judgement when get max latency

Reviewed-by: xie xingguo <xie.xingguo@zte.com.cn>
Reviewed-by: Pan Liu <liupan1111@gmail.com>

Merge pull request #15999 from Yan-waller/wip-waller-0629message

msg/MOSDOpReply: fix missing trace decode

Reviewed-by: Kefu Chai <kchai@redhat.com>
Reviewed-by: Casey Bodley <cbodley@redhat.com>

Merge pull request #16037 from gmayyyha/monmapmonitor-use-func

mon/MonmapMonitor: use __func__ instead of hard code function name

Reviewed By: Joao Eduardo Luis <jecluis@gmail.com>

do all valgrind runs on centos

We are fighting two issues with valgrind on ubuntu (xenial, yakkety,
and z):

http://tracker.ceph.com/issues/18126
http://tracker.ceph.com/issues/20360

Revert this when it is fixed.

Signed-off-by: Sage Weil <sage@redhat.com>

qa: stop testing btrfs

This is too noisy and we don't recommend it anyway.

See http://tracker.ceph.com/issues/20169

Signed-off-by: Sage Weil <sage@redhat.com>

PendingReleaseNotes: add a note on choose_args encoding

It's unlikely that it's been used yet, but worth a note.

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

crush/CrushWrapper: crush_choose_arg::ids should be __s32

crush_choose_arg::ids array is encoded on the wire.  int is not fixed
size -- use __s32 instead (crush_bucket::id is __s32).

This was introduced in commit dbe36e08be00 ("crush: compile/decompile
crush_choose_arg_map") under SERVER_LUMINOUS bit.

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

crush/CrushWrapper: encode choose_args.size() as __u32

choose_args.size() returns size_t, which is not fixed size.  Given that
choose_args code just open-codes our std::map encoder, do what we do for
std::map.

This was introduced in commit dbe36e08be00 ("crush: compile/decompile
crush_choose_arg_map") under SERVER_LUMINOUS bit.

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

mon/MonmapMonitor: use __func__ instead of hard code function name

Signed-off-by: Yanhu Cao <gmayyyha@gmail.com>

systemd: Add explicit Before=ceph.target

The PartOf= and WantedBy= directives in the various systemd
unit files and targets create the following logical hierarchy:

- ceph.target
  - ceph-fuse.target
    - ceph-fuse@.service
  - ceph-mds.target
    - ceph-mds@.service
  - ceph-mgr.target
    - ceph-mgr@.service
  - ceph-mon.target
    - ceph-mon@.service
  - ceph-osd.target
    - ceph-osd@.service
  - ceph-radosgw.target
    - ceph-radosgw@.service
  - ceph-rbd-mirror.target
    - ceph-rbd-mirror@.service

Additionally, the ceph-{fuse,mds,mon,osd,radosgw,rbd-mirror}
targets have WantedBy=multi-user.target.  This gives the
following behaviour:

- `systemctl {start,stop,restart}` of any target will restart
  all dependent services (e.g.: `systemctl restart ceph.target`
  will restart all services; `systemctl restart ceph-mon.target`
  will restart all the mons, and so forth).
- `systemctl {enable,disable}` for the second level targets
  (ceph-mon.target etc.) will cause depenent services to come
  up on boot, or not (of course the individual services can
  be enabled or disabled as well - for a service to start
  on boot, both the service and its target must be enabled;
  disabling either will cause the service to be disabled).
- `systemctl {enable,disable} ceph.target` has no effect on
  whether or not services come up at boot; if the second level
  targets and services are enabled, they'll start regardless of
  whether ceph.target is enabled.  This is due to the second
  level targets all having WantedBy=multi-user.target.
- The OSDs will always start regardless of ceph-osd.target
  (unless they are explicitly masked), thanks to udev magic.

So far, so good.  Except, several users have encountered
services not starting with the following error:

  Failed to start ceph-osd@5.service: Transaction order is
  cyclic. See system logs for details.

I've not been able to reproduce this myself in such a way as to
cause OSDs to fail to start, but I *have* managed to get systemd
into that same confused state, as follows:

- Disable ceph.target, ceph-mon.target, ceph-osd.target,
  ceph-mon@$(hostname).service and all ceph-osd instances.
- Re-enable all of the above.

At this point, everything is fine, but if I then subseqently
disable ceph.target, *then* try `systemctl restart ceph.target`,
I get "Failed to restart ceph.target: Transaction order is cyclic.
See system logs for details."

Explicitly adding Before=ceph.target to each second level target
prevents systemd from becoming confused in this situation.

Signed-off-by: Tim Serong <tserong@suse.com>

os/bluestore/KernelDevice: make the amount of reaped aio tunable

Signed-off-by: Haodong Tang <haodong.tang@intel.com>

os/bluestore/KernelDevice: batch aio submit

reduce unnessary system call(io_submit).

Signed-off-by: Haodong Tang <haodong.tang@intel.com>

msg/async: increase worker reference with local listen table enabled backend

Fixes: http://tracker.ceph.com/issues/20390
Signed-off-by: Haomai Wang <haomai@xsky.com>

mon,mgr: track available modules in MgrMap too

Require --force to enable a module that isn't reported as available.

Signed-off-by: Sage Weil <sage@redhat.com>

doc: update docs for enabling mgr modules

Signed-off-by: Sage Weil <sage@redhat.com>

mon,mgr: manage mgr module list in mgrmap

Kill old mgr_modules option.

Add new mgr_initial_modules option, on the mon, for the initial cluster
mgrmap.

Add ls, enable, disable commands.

Respawn mgr if the module list changes.  In the future we could enable
new modules without a full restart, but disabling probably requires (and
is best handled by) a respawn.

Signed-off-by: Sage Weil <sage@redhat.com>