doc/cephadm: Add iscsi

Fixes https://tracker.ceph.com/issues/49488

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit 46a203a5c6f4552fd4618e659e8ac7abbba0f186)

doc/cephadm: group NFS sections into one chapter

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit 8d07e35b2f10f8b6b6bb039e6c066ff3d1056d86)

doc/cephadm: rename monitoring chapter title

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit 769642c77d83e1c70b61e064151e92498f92c8a5)

doc/cephadm: group MON sections into one chapter

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit b3846f6f6a4ff3daf7c26eb087992bd230ee9a47)

doc/cephadm: make custom containers its own chapter

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit 0fb2ccb49b765d589f56b355589dc811fbf66d63)

doc/cephadm: group RGW mgmt sections into one chapter

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit 818ed4f2025d46a0822d025c5cae5eeb161099c3)

doc/cephadm: move scheduler topic to service mgmt

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit 4f3305f3bb44d9169ef74b10cb9b95e14f177137)

Conflicts:
doc/cephadm/index.rst

doc/cephadm: move unmanaged=true to service-mgmt.rst

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit 31470cfd5c04011dc69c106a921fc4dbf561ce46)

doc/cephadm: group general service mgmt sections into one chapter

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit 11fe5ef7cf7fb93440d58fee5594fad8bd7ef32b)

doc/cephadm: group OSD mgmt sections into one chapter

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit 74bcd7c5f6fe1c6ddb1b01bae8e964d71b57d5a1)

doc/cephadm: Move FQDN chapter to host mgmt.rst

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit a3036d974c8c25eda6c50e6f997c826d4b8cea02)

doc/cephadm: Move SSH config from operations to host-mgmt.rst

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit 1dbac10c710ee92ed41415d2b30af58a930eecf9)

doc/cephadm: group host mgmt sections into one chapter

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit 88e8e91f7289b4c47056ecae1d2d198c7e7259a8)

cephadm: fix bug in orphan-initial-daemons logic

Signed-off-by: Daniel Pivonka <dpivonka@redhat.com>
(cherry picked from commit 0197c7b13290e1d9b2a719ed6f2f70987c9592e3)

mgr/orch: drop __all__ from __init__.py

`__all__` is only required if you want to modify the `form .. import *`
behavior. We don't need it.

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit b8dbdb6530ce3ce98df42abc3825e0876cf8388a)

mgr/cephadm: add DaemonDescriptionStatus

Signed-off-by: Daniel Pivonka <dpivonka@redhat.com>
Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit ae83ffc7897a0912308001f55876fd51dd56ad40)

cephadm: version command hide traceback when login is needed

if image comes from a authenticated registry and login has not been done yet the traceback should be hidden and only show the error message

Signed-off-by: Daniel Pivonka <dpivonka@redhat.com>
(cherry picked from commit 4c1c656d009230894674d33efed9c31b2bc2581d)

doc/cephadm: troubleshooting: manually deploy MGR

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit a4179acbce3b9e71abb526a5b46436d6a3b96c82)

cephadm: fix port_in_use when IPv6 is disabled

Do not return "port is in use" when the protocol family tested is not
supported (due to being deactivated).

Fixes: https://tracker.ceph.com/issues/49273
Signed-off-by: Patrick Seidensal <pseidensal@suse.com>
(cherry picked from commit a0ffcec4af7a6f464439023d3dd0ccb62c8a856c)

cephadm: Allow to use paths in all <_devices> drivegroup sections

This will allow to use specific device paths in any osd component.

Signed-off-by: Juan Miguel Olmo Martínez <jolmomar@redhat.com>
(cherry picked from commit d02683efbfcd5aa1437ebb9e73f89db38f7e7aa6)

mgr/cephadm: error if service action called with daemonless service

Fixes: https://tracker.ceph.com/issues/49348
Signed-off-by: Adam King <adking@redhat.com>
(cherry picked from commit 0618c69dc5255bd407595250c6980cf4929a8ce6)

mgr/cephadm: fix up the strings reporting osd ids

Signed-off-by: Sage Weil <sage@newdream.net>
(cherry picked from commit a1ff3a9952778c1f20836b806de9fa5606432137)

mgr/cephadm: remove daemon before osd destroy/purge

Otherwise it doesn't work!

Drop the fullname property: it is always "osd.{self.osd_id}".

Signed-off-by: Sage Weil <sage@newdream.net>
(cherry picked from commit b5eab0ddfa0bb8ae7b1a6aec4ea2e4257a01a045)

mgr/cephadm: simplify OSD __str__ for drain

Signed-off-by: Sage Weil <sage@newdream.net>
(cherry picked from commit ca4050b057296d3c62deefca0ffcb4f640b30102)

mgr/cephadm: make drain adjust crush weight if not replacing

If we are replacing an OSD, we should mark it out and then back in
again when a new device shows up.  However, if we are going to
destroy an OSD, we should just weight it to 0 in crush, so that data
doesn't move again once the OSD is purged.

Signed-off-by: Sage Weil <sage@newdream.net>
(cherry picked from commit 4fc1309f281356db0a074da22aa6f2daa034df8d)

mgr/cephadm: less log noise from osd drain code

Signed-off-by: Sage Weil <sage@newdream.net>
(cherry picked from commit e2f0e56ddf3197f220c5a43c79d6bc43e4b135ce)

mgr/cephadm: fix 'orch daemon add osd ...'

When adding an osd daemon explicitly, there is no created timestamp
for the spec, and we should never not apply it.

Fixes: b129c1312113f56a227caeb535f656f5a090a85f
Signed-off-by: Sage Weil <sage@newdream.net>
(cherry picked from commit e8643275e5d92af9539e60a7a80ef13d0f27af64)

mgr/cephadm/upgrade: fix typo

Signed-off-by: Sage Weil <sage@newdream.net>
(cherry picked from commit 8e05d9759b607582deaed872d4e4d61c1cc30689)

mgr/cephadm: remove spec from CephadmDaemonDeploySpec

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit d8bcaac10a1a7cdc8689375f37f5d59d86148314)

mgr/cephadm/upgrade: restart mgr after mons upgrade to pacific

If we are upgrading to pacific, we need to restart the mgr *after* the
mons have a pacific quorum so that they can get the kv subscriptions.

Signed-off-by: Sage Weil <sage@newdream.net>
(cherry picked from commit 4d7dd0012962202532bde4fbc47554186c95b9fb)

mgr/cephadm: use get_foreign_ceph_option() instead of 'config get' mon command

Signed-off-by: Sage Weil <sage@newdream.net>
(cherry picked from commit 9b5cbf40f8328060d7c6d04a474ee87d5bf03ca3)

Merge pull request #39741 from tchaikov/pacific-49502

pacific: mgr/rbd_support: bail out if snapshot mirroring is not enabled

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

Merge PR #39856 into pacific

* refs/pull/39856/head:
qa/distro/ubuntu_20.04_podman: Avoid getting asked
qa/suites/rados/cephadm: drop centos/rhel cephadm tests for the moment
qa/sites/rados/cephadm/thrash: rename 3-tasks.yaml/ -> 3-tasks/
qa/suites/rados/cephadm: adjust distros
qa/suites/upgrade: use kubic; test all distros
qa/suites/rados/cephadm/upgrade: use kubic on centos
qa: new kubic distro files; use kubic podman for centos/rhel
qa/suites/rados/cephadm: Add 20.04 podman:testing

Reviewed-by: Sebastian Wagner <swagner@suse.com>

qa/distro/ubuntu_20.04_podman: Avoid getting asked

... for confic changes interactively

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit 0defae4e65dee123ee0b8474bf3cb417ddff7eb1)

Merge PR #39628 into pacific

* refs/pull/39628/head:
mgr/dashboard: set security headers

Reviewed-by: Nizamudeen A <nia@redhat.com>

Merge PR #39702 into pacific

* refs/pull/39702/head:
mon: simplify fs 'services' view of 'ceph status'
mds: add volumes + status to data: section of 'ceph status'

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

Merge PR #39727 into pacific

* refs/pull/39727/head:
mgr/telemetry: check if 'ident' channel is active
qa/tasks/mgr/dashboard: identify cluster in telemetry test

Reviewed-by: Yaarit Hatuka <yaarithatuka@gmail.com>

Merge PR #39788 into pacific

* refs/pull/39788/head:
mon/ConfigMonitor: make config changes via KVMonitor's pending set

Reviewed-by: Neha Ojha <nojha@redhat.com>

Merge PR #39805 into pacific

* refs/pull/39805/head:
src/global/signal_handler.h: fix preprocessor logic for alpine

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

Merge PR #39811 into pacific

* refs/pull/39811/head:
doc/rbd: clarify QCOW2 migration support

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

Merge pull request #39806 from liewegas/pr-39665-pacific

pacific: qa/rados/dashbord use "random" objectore

Reviewed-by: Neha Ojha <nojha@redhat.com>
Reviewed-by: Sebastian Wagner <sebastian.wagner@suse.com>

Merge pull request #39796 from dzafman/wip-48984-pacific

pacific: osd: Don't clear m_planned_scrub until scrub finishes successfully

Reviewed-by: Neha Ojha <nojha@redhat.com>

Merge pull request #39773 from myoungwon/pacific-backport-49482

pacific: osd, test: fix to avoid fail notification when testing manifest refcount

Reviewed-by: Neha Ojha <nojha@redhat.com>
Reviewed-by: Samuel Just <sjust@redhat.com>

Merge pull request #39753 from ifed01/wip-ifed-bluefs-zero-read-retry-pac

pacific: os/bluestore: Add option to check BlueFS reads

Reviewed-by: Adam Kupczyk <akupczyk@redhat.com>

Merge pull request #39748 from liewegas/pr-39631-pacific

pacific: mon/OSDMonitor: do not mark newly created OSDs OUT

Reviewed-by: Neha Ojha <nojha@redhat.com>

Merge pull request #39688 from ifed01/wip-ifed-fix-huge-bluefs-writes-pac

pacific: os/bluestore: fix huge(>4GB) writes from RocksDB to BlueFS.

Reviewed-by: Neha Ojha <nojha@redhat.com>

qa/suites/rados/cephadm: drop centos/rhel cephadm tests for the moment

...until we sort out the podman problems!  :( :(

Signed-off-by: Sage Weil <sage@newdream.net>
(cherry picked from commit e42223c4fc8ca42e03a5721ae5b0bc5362f144ce)

qa/sites/rados/cephadm/thrash: rename 3-tasks.yaml/ -> 3-tasks/

Signed-off-by: Sage Weil <sage@newdream.net>
(cherry picked from commit 5f5472b7daede39a7bf35bf9135fc4fc75ac5a86)

qa/suites/rados/cephadm: adjust distros

Signed-off-by: Sage Weil <sage@newdream.net>
(cherry picked from commit 93e77ac3c81b5e6fc363786429b7eee0ac8567c3)

qa/suites/upgrade: use kubic; test all distros

Signed-off-by: Sage Weil <sage@newdream.net>
(cherry picked from commit b378c042296858ec4d1c63a42cd257580ca11c96)

qa/suites/rados/cephadm/upgrade: use kubic on centos

Signed-off-by: Sage Weil <sage@newdream.net>
(cherry picked from commit f4de8738ca86b7b31998718b53c8b7659fcdd27f)

qa: new kubic distro files; use kubic podman for centos/rhel

The current centos/rhel version of podman (2.2.1) is broken.

- create new qa/distros/podman/* files that install kubic podman
- include centos/rhel variants
- adjust cephadm jobs to use new yaml files
- remove old qa/distros/all/*_podman.yaml files

Signed-off-by: Sage Weil <sage@newdream.net>
(cherry picked from commit 11beb7287084b56ac7cbd45ac3d28b3d9fa4cc5b)

qa/suites/rados/cephadm: Add 20.04 podman:testing

make sure next podman versions aren't breaking cephadm

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit 2216350b567102db4ab80dc42dfff9dced2b9740)

Merge pull request #39808 from liewegas/pr-39801-pacific

pacific: mgr/dashboard: fix issues related with PyJWT versions >=2.0.0

Reviewed-by: Laura Paduano <lpaduano@suse.com>
Reviewed-by: Ernesto Puerta <epuertat@redhat.com>

Merge PR #39737 into pacific

* refs/pull/39737/head:
mgr/DaemonServer: osd ok-to-stop: return json when there are unknown PGs
doc/man/8/ceph: document --max option
src/test/osd/safe-to-destroy: adjust test
ceph: print command output to stdout even on error
mgr/DaemonServer: include details in 'osd ok-to-stop' output
mgr: add --max <n> to 'osd ok-to-stop' command
mgr: relax osd ok-to-stop condition on degraded pgs

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

Merge pull request #39517 from k0ste/wip-49325-pacific

pacific: mgr/dashboard: fix MTU Mismatch alert

Reviewed-by: Aashish Sharma <aasharma@redhat.com>
Reviewed-by: Avan Thakkar <athakkar@redhat.com>
Reviewed-by: Nizamudeen A <nia@redhat.com>

Merge PR #39736 into pacific

* refs/pull/39736/head:
crush/CrushWrapper: rebuild shadow tree on 'osd crush reweight-subtree'
crush/CrushWrapper: update shadow trees on update_item()

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

Merge PR #39734 into pacific

* refs/pull/39734/head:
mon/PGMap: include the per-pool usage breakdown in pg dump

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

Merge PR #39735 into pacific

* refs/pull/39735/head:
mon/OSDMonitor: fix safety/idempotency of {set,rm}-device-class

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

Merge PR #39597 into pacific

* refs/pull/39597/head:
osd: Don't clear m_planned_scrub until scrub finishes successfully

Reviewed-by: Brad Hubbard <bhubbard@redhat.com>

mgr/DaemonServer: osd ok-to-stop: return json when there are unknown PGs

In 791952cc01201010f298033003ba52374cc0159f we switched to return JSON
both on success and fail to describe which PGs are affected or are blocking
the ability to stop/restart OSDs.  Do the same for the case where
some PG states are unknown (i.e., just after a mgr restart) so that
the cephadm upgrade process can unconditionally expect a JSON result.

Signed-off-by: Sage Weil <sage@newdream.net>
(cherry picked from commit 2cce16537c9f9e92bf0a8fce746d0364cbb2d15d)

Merge pull request #39524 from rhcs-dashboard/wip-49326-pacific

pacific: mgr/dashboard: avoid using document.write()

Reviewed-by: Laura Paduano <lpaduano@suse.com>
Reviewed-by: Nizamudeen A <nia@redhat.com>

Merge pull request #39647 from rhcs-dashboard/wip-49390-pacific

pacific: mgr/dashboard: fix tooltip for Provisioned/Total Provisioned fields

Reviewed-by: Laura Paduano <lpaduano@suse.com>
Reviewed-by: Nizamudeen A <nia@redhat.com>

Merge pull request #39816 from cbodley/wip-49585

pacific: valgrind: suppress operator delete[] for tcmalloc

Reviewed-by: Daniel Gryniewicz <dang@redhat.com>

valgrind: suppress operator delete[] for tcmalloc

suppresses a mismatched operator delete[] from the librados client's
destructor:

<error>
  <unique>0xeef7eb</unique>
  <tid>1</tid>
  <kind>MismatchedFree</kind>
  <what>Mismatched free() / delete / delete []</what>
  <stack>
    <frame>
      <ip>0x4C32EA0</ip>
      <obj>/usr/libexec/valgrind/vgpreload_memcheck-amd64-linux.so</obj>
      <fn>operator delete[](void*, unsigned long)</fn>
      <dir>/builddir/build/BUILD/valgrind-3.16.0/coregrind/m_replacemalloc</dir>
      <file>vg_replace_malloc.c</file>
      <line>660</line>
    </frame>
    <frame>
      <ip>0x5FBEB86</ip>
      <obj>/usr/lib64/librados.so.2.0.0</obj>
    </frame>
    <frame>
      <ip>0x5FBEF2C</ip>
      <obj>/usr/lib64/librados.so.2.0.0</obj>
    </frame>
    <frame>
      <ip>0x5F7F7C5</ip>
      <obj>/usr/lib64/librados.so.2.0.0</obj>
      <fn>librados::v14_2_0::Rados::shutdown()</fn>
    </frame>

Fixes: https://tracker.ceph.com/issues/49501
Signed-off-by: Casey Bodley <cbodley@redhat.com>
(cherry picked from commit bf087c3f840216a704600bb2436c643159421ea2)

mgr/dashboard: fix issues related with PyJWT versions >=2.0.0

Fixes: https://tracker.ceph.com/issues/49574
Signed-off-by: Alfonso Martínez <almartin@redhat.com>
(cherry picked from commit 4b96bb51e8f133badd7bc651bcb4dcd755b43d75)

doc/rbd: clarify QCOW2 migration support

Signed-off-by: Jason Dillaman <dillaman@redhat.com>
(cherry picked from commit 68fda590cf88e6b232dfb573236fe3ae3949db28)

Merge pull request #39568 from tchaikov/pr-39365-pacific

pacific: mgr/devicehaelth: extract and present SSD wear

Reviewed-by: Alfonso Martínez <almartin@redhat.com>
Reviewed-by: Nizamudeen A <nia@redhat.com>
Reviewed-by: Sage Weil <sage@redhat.com>

qa/rados/dashbord use "random" objectore

1. removed symlink to `.qa/objectstore`
2. added `random-objectstore&` to `.qa/objectstore`

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit 72229a098b878036e2738147814898bc65efcaea)

src/global/signal_handler.h: fix preprocessor logic for alpine
Fixes: https://tracker.ceph.com/issues/49494
Co-authored-by: Kefu Chai <tchaikov@gmail.com>
Signed-off-by: Duncan Bellamy <dunk@denkimushi.com>
(cherry picked from commit 387f33e7265825ae4cdd3580289f5154e250b05b)

Merge PR #39525 into pacific

* refs/pull/39525/head:
ceph-volume: pass --filter-for-batch from drive-group subcommand

Reviewed-by: Rishabh Dave <ridave@redhat.com>

Merge PR #39468 into pacific

* refs/pull/39468/head:
Update batch.py

Reviewed-by: Rishabh Dave <ridave@redhat.com>

qa/tasks/mgr/dashboard: add wear_level to devices schema

Signed-off-by: Sage Weil <sage@newdream.net>
(cherry picked from commit 6f6792cf51f9042452bc79df30b3a4d02dd5ef96)

mgr/devicehealth: extract+store wear level from metrics scraping

When we scrape and store health metrics for a device, extract the wear
level from the JSON.  If present, also store it in the config-key
per-device metadata.

Signed-off-by: Sage Weil <sage@newdream.net>
(cherry picked from commit 4840507cfcdd5182003671994d0bc9604d072e3e)

 Conflicts:
src/pybind/mgr/devicehealth/module.py
 - type annotations for _get_device_metrics

mgr/DaemonServer: show wear level in 'device ls' output

Signed-off-by: Sage Weil <sage@newdream.net>
(cherry picked from commit 8f93e3b55351df0bc8308e0cace482aabb15e8c8)

mgr/DaemonState: store optional wear_level for devices

Signed-off-by: Sage Weil <sage@newdream.net>
(cherry picked from commit c52fcb51afb6da2dccd6954a40f0f1c616324855)

common/blkdev: collect non-SMART data too

Call smartctl with -x instead of -a:

       -a, --all
              Prints  all SMART information about the disk, or TapeAlert infor‐
              mation about the tape drive or changer.  For ATA devices this  is
              equivalent to
              '-H -i -c -A -l error -l selftest -l selective'
              and for SCSI, this is equivalent to
              '-H -i -A -l error -l selftest'.
              For NVMe, this is equivalent to
              '-H -i -c -A -l error'.
              Note  that  for  ATA  disks  this  does  not enable the non-SMART
              options and the SMART options which require  support  for  48-bit
              ATA commands.

vs

       -x, --xall
              Prints all SMART and non-SMART information about the device.  For
              ATA devices this is equivalent to
              '-H -i -g all -g wcreorder -c -A  -f  brief  -l  xerror,error  -l
              xselftest,selftest -l selective -l directory -l scttemp -l scterc
              -l devstat -l defects -l sataphy'.
              and for SCSI, this is equivalent to
              '-H -i -g all -A -l error -l selftest -l background -l sasphy'.
              For NVMe, this is equivalent to
              '-H -i -c -A -l error'.

Signed-off-by: Sage Weil <sage@newdream.net>
(cherry picked from commit 176bb54da90e404646c0e80e0c3ea804bb8a58d6)

mgr/DaemonState: fix serialization/storage of device life expectancy min/max

These weren't encoded as strings, which means they were effectively
forgotton on each mgr restart.  Sigh.

Fixes: https://tracker.ceph.com/issues/49215
Signed-off-by: Sage Weil <sage@newdream.net>
(cherry picked from commit 78b72b0be752fce855203b28b558133a0d47eba8)

pybind/mgr: silence flake8 warnings

silence following warnings from flake8

* line too long (107 > 100 characters) [E501]

Signed-off-by: Kefu Chai <kchai@redhat.com>
(cherry picked from commit 367f264100778717890b7fb69668e44f0a22d8ce)

pybind/mgr/mgr_module: correct annotation

Signed-off-by: Kefu Chai <kchai@redhat.com>
(cherry picked from commit 19cdc4325ce47bef951fdab41fa15e6acadea1f7)

mgr/devicehealth: remove handle_command()

as commands are collected by COMMANDS already.

Signed-off-by: Kefu Chai <kchai@redhat.com>
(cherry picked from commit 1361b3060e7643a3a0f2f59b72eb5d9201c21bcd)

mgr/devicehealth: return -1 if prediction plugin is not available

Signed-off-by: Kefu Chai <kchai@redhat.com>
(cherry picked from commit bb611c64ee5120272215683b4b5ac4397f9c5ea6)

mgr/devicehealth: add more annotations

Signed-off-by: Kefu Chai <kchai@redhat.com>
(cherry picked from commit 48438cccedb383e19cab743dd0222d6ec6265588)

osd: Don't clear m_planned_scrub until scrub finishes successfully

Fixes: https://tracker.ceph.com/issues/48984
Signed-off-by: David Zafman <dzafman@redhat.com>
(cherry picked from commit 0b63c850a251f75bcb7436efefe4f24480c4ed7e)

Merge pull request #39769 from ivancich/wip-49185-pacific

pacific: rgw: tooling to locate rgw objects with missing rados components

Reviewed-by: Michael Kidd <linuxkidd@gmail.com>

Merge pull request #39764 from ivancich/wip-49194-pacific

pacific: rgw: allow rgw-orphan-list to handle intermediate files w/ binary data

Reviewed-by: Thomas Serlin <tserlin@redhat.com>

Merge pull request #39675 from dillaman/wip-librbd-backports-pacific-6

pacific: librbd: miscellaneous backports

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

Merge pull request #39635 from vshankar/wip-49425

pacific: cephfs-mirror: run mirror daemon with valgrind

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

Merge pull request #39508 from k0ste/wip-49315-pacific

pacific: mgr/dashboard: prometheus alerting: add some leeway for package drops and errors

Reviewed-by: Avan Thakkar <athakkar@redhat.com>

Merge pull request #39466 from petrutlucian94/wip-49250-pacific

pacific: common: Add Windows ETW support

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

mon/ConfigMonitor: make config changes via KVMonitor's pending set

We need to ensure that changes we make to the kv store (config/...)
are proposed via KVMonitor so that they are properly versioned there
and shared with subscribers (notably, the mgr).

Fixes: bb7ebc41532aeb23cff2241ab07b3f01c2f57ddd
Signed-off-by: Sage Weil <sage@newdream.net>
(cherry picked from commit dab72abd0ae8a3038f73dbe0983b2eaef3937ef6)

Merge pull request #39752 from rhcs-dashboard/wip-49546-pacific

pacific: mgr/dashboard: Splitting tenant$user when creating rgw user

Reviewed-by: Aashish Sharma <aasharma@redhat.com>
Reviewed-by: Alfonso Martínez <almartin@redhat.com>
Reviewed-by: Avan Thakkar <athakkar@redhat.com>

osd, test: wait if the snapshot is deleting

After calling selfmanaged_snap_remove, we don't know
when trimming snapshot is finished.
So, we make the OSD to return EBUSY if the snapshot in removed_snap_queue,
then the unit test waits the completion

Signed-off-by: Myoungwon Oh <myoungwon.oh@samsung.com>
(cherry-picked from commit d6f9f23012b9ac133e767bd07a804707ba2207ef)

src/test: fix to avoid fail notification when testing manifest refcount

Due to false-positive design on manifest snap refcounting,
a message to decrement the refcount can be missing.
This commit checks whether the manifest object's state is correct
when such mismatch happens to prevent aborting unit test.

Signed-off-by: Myoungwon Oh <myoungwon.oh@samsung.com>
(cherry-picked from commit d0369dc73ee9a911e39642f3b63c4d0f17b04ce3)

osd: remove unnecessary check

Signed-off-by: Myoungwon Oh <myoungwon.oh@samsung.com>
(cherry-picked from commit 1109a59e2941cf4b16798cd04941daf350ec0d22)

osd, test: return the number of references of chunked object

Signed-off-by: Myoungwon Oh <myoungwon.oh@samsung.com>
(cherry-picked from commit cb7cf5ba545386e1bf010d655f0736d0ef5a29ae)

test: add cls_has_chunk test to count chunk in snapshot

Signed-off-by: Myoungwon Oh <myoungwon.oh@samsung.com>
(cherry-picked from commit 11b52b4d00c7d56f0a571fd9208740932b0d00dd)

osd: add has_manifest_chunk() to check chunks in snapshot

cls_has_chunk does not cover snapshotted manifest object.
This leads to unexpected behavior during chunk scrub.

Signed-off-by: Myoungwon Oh <myoungwon.oh@samsung.com>
(cherry-picked from commit bdcedd4e9f2efed59b1e5c2d40171e6da595a259)

rgw: add rgw-gap-list-comparator tool

The rgw-gap-list tool can produce a number of false positives when the
cluster is being used during its run. One technique to minimize the
number of false positives is to run the tool twice and look for the
objects that appear in both lists. The rgw-gap-list-comparator tool is
designed to do this comparison.

Signed-off-by: Michael Kidd <linuxkidd@gmail.com>
(cherry picked from commit 3c5154f010a34f9845b40e49a7870771e7f5372e)

rgw: add rgw-gap-list tool

Due to a prior bug (pr: 38228) tail rados objects of some RGW objects
could have been incorrectly deleted. This tool is designed to look for
such cases. It essentially does the opposite of rgw-orphan-list,
looking for rados objects that RGW expects to be there, but which are
not to be found.

IMPORTANT: This is very experimental at this point in time, and any
"results" produced should be verified by other means.

Signed-off-by: J. Eric Ivancich <ivancich@redhat.com>
Signed-off-by: Michael Kidd <linuxkidd@gmail.com>
(cherry picked from commit 07b42195fbbcd27e330cb1daa35e77e0952f8a3c)