doc: Move ceph-deploy docs to doc/install/ceph-deploy

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>

Merge pull request #33867 from dillaman/wip-43590

librbd: optimize image copy state machine to use fast-diff

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

Merge PR #33930 into octopus

* refs/pull/33930/head:
mgr/cephadm: implement pause/resume to suspect non-monitoring background work
mgr/orch: add pause/resume hooks

Reviewed-by: Michael Fritch <mfritch@suse.com>

Merge PR #33868 into octopus

* refs/pull/33868/head:
mgr/rook: implement remove_daemons()
mgr/rook: fix mds/fs create/apply/update
mgr/rook: use ceph_daemon_id label consistently
mgr/rook: show filesystems in 'orch ls'
mgr/orch: 'age' -> 'ago' in 'orch ls' output
mgr/rook: initial pass at 'orch ls'
mgr/rook: include timestamps in 'orch ps'
mgr/orch: shorten container_id (hash) in cephadm, not orch ps
mgr/rook: include container_image_name in 'orch ps'
mgr/rook: fix list_daemons host arg

Reviewed-by: Juan Miguel Olmo <jolmomar@redhat.com>
Reviewed-by: Sebastian Wagner <swagner@suse.com>

Merge PR #33932 into octopus

* refs/pull/33932/head:
cephadm: fix datetime regexp to capture at most 6 digits

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

Merge pull request #33900 from dillaman/wip-44459

qa/suites/rbd: removed OpenStack tempest test cases

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

Merge pull request #33840 from trociny/wip-schedule-optimize-scan

mgr/rbd_support: don't scan pools that don't have schedules

Reviewed-by: Jason Dillaman <dillaman@redhat.com>

Merge PR #33914 into octopus

* refs/pull/33914/head:
mgr/orch: Cleanup python imports of the interface
mgr/orch: Properly handle NotImplementedError
pybind/mgr: Rm PersistentStoreDict

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

mgr/rook: implement remove_daemons()

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

mgr/rook: fix mds/fs create/apply/update

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

mgr/rook: use ceph_daemon_id label consistently

In particular, for mds's, it's $fsname-$rank for mds.$fsname-$rank,
which is what we want.

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

mgr/rook: show filesystems in 'orch ls'

The count vs size discrepancy is weird since rook things in terms of
ranks (*2 if standby) vs just a count.

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

mgr/orch: 'age' -> 'ago' in 'orch ls' output

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

mgr/rook: initial pass at 'orch ls'

This doesn't cover mds.* or rgw.* yet.

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

mgr/rook: include timestamps in 'orch ps'

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

mgr/orch: shorten container_id (hash) in cephadm, not orch ps

This lets us present a longer, human-readable id for rook.

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

mgr/rook: include container_image_name in 'orch ps'

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

mgr/rook: fix list_daemons host arg

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

cephadm: fix datetime regexp to capture at most 6 digits

trailing `0` digits are not shown in the precision

Signed-off-by: Michael Fritch <mfritch@suse.com>

Merge pull request #33821 from votdev/issue_43446_bucket_lock

mgr/dashboard: Create bucket with x-amz-bucket-object-lock-enabled

Reviewed-by: Alfonso Martínez <almartin@redhat.com>
Reviewed-by: Stephan Müller <smueller@suse.com>

Merge pull request #31997 from bk201/wip-43062

mgr/dashboard: support removing OSDs in OSDs page

Reviewed-by: Ernesto Puerta <epuertat@redhat.com>
Reviewed-by: Joshua Schmid <jschmid@suse.de>
Reviewed-by: Laura Paduano <lpaduano@suse.com>
Reviewed-by: Sebastian Wagner <swagner@suse.com>
Reviewed-by: Tiago Melo <tmelo@suse.com>
Reviewed-by: Volker Theile <vtheile@suse.com>

Merge pull request #33920 from LenzGr/update-404-image

mgr/dashboard: Updated octopus image on 404 page

Reviewed-by: Tatjana Dehler <tdehler@suse.com>
Reviewed-by: Volker Theile <vtheile@suse.com>

mgr/cephadm: implement pause/resume to suspect non-monitoring background work

If the user does 'orch pause', suspend all background work that makes
actual changes.

Continue to do read-only operations, like checking host connectivity
and scraping daemon and device status.

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

Merge pull request #33858 from tspmelo/wip-use-cd-modal

mgr/dashboard: Use ModalComponent in all modals

Reviewed-by: Ernesto Puertat <epuertat@redhat.com>
Reviewed-by: Stephan Müller <smueller@suse.com>

mgr/orch: add pause/resume hooks

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

Merge PR #33926 into octopus

* refs/pull/33926/head:
python: fixate mypy requirement to mypy==0.770

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

Merge PR #33874 into octopus

* refs/pull/33874/head:
qa/suites/powercycle: install build deps for building xfstest

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

Merge PR #33813 into octopus

* refs/pull/33813/head:
osd/PGStateUtils: initialize NamedState::enter_time.
osd/PGStatueUtils: remove unused member variable

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

Merge PR #33814 into octopus

* refs/pull/33814/head:
osd/OSDMap: stop encoding osd_state with >8 bits wide states only for old client

Reviewed-by: Sage Weil <sage@redhat.com>
Reviewed-by: yanjun <yan.jun8@zte.com.cn>

Merge PR #33064 into octopus

* refs/pull/33064/head:
cephadm: add version to `command_ls` output
cephadm: add type checking to `update_filewalld`
cephadm: allow prepare-host to start an enabled service
cephadm: add type checking for `check_host` and `prepare_host`
cephadm: generalize logic for checking and enabling units
cephadm: add 'CEPH_CONF' to the NFS ganesha container envs
cephadm: trim nfs.json sample
qa/workunits/cephadm/test_cephadm.sh: systemctl stop nfs-server
qa/workunits/cephadm/test_cephadm.sh: make pgs available
cephadm: add some log lines
cephadm: check port in use
cephadm: add/remove nfs ganesha grace
cephadm: update firewalld with nfs service
qa/workunits/cephadm/test_cephadm.sh: add nfs-ganesha test
cephadm: add ganasha.conf
cephadm: add NFSGanesha deployment type
cephadm: consolidate list of supported daemons
cephadm: use keyword instead of positional args

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

Merge PR #33855 into octopus

* refs/pull/33855/head:
mgr/cephadm: reconfigure ceph daemons when monmap changes

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

Merge PR #33829 into octopus

* refs/pull/33829/head:
cephadm: log when we infer the image
cephadm: Avoid `inspect` call by using ceph label
cephadm: Infer ceph image

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

Merge pull request #33919 from LenzGr/octopus-documentation

documentation: Updated Dashboard Features, improved flow

Reviewed-by: Kiefer Chang <kiefer.chang@suse.com>
Reviewed-by: Tatjana Dehler <tdehler@suse.com>
Reviewed-by: Volker Theile <vtheile@suse.com>

Merge PR #33871 into octopus

* refs/pull/33871/head:
qa/suites/rados/verify: debug_ms = 1, osd_heartbeat_grace = 60

Reviewed-by: xie xingguo <xie.xingguo@zte.com.cn>
Reviewed-by: Brad Hubbard <bhubbard@redhat.com>

Merge PR #33890 into octopus

* refs/pull/33890/head:
doc/install: rethink docs install page (similar to PR#33613)
doc/install: Rethink doc/install

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

python: fixate mypy requirement to mypy==0.770

I don't like suprises when upstream publishes a new mypy version.

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>

mgr/dashboard: Append "Modal" to all modal components names

Fixes: https://tracker.ceph.com/issues/44547
Signed-off-by: Tiago Melo <tmelo@suse.com>

mgr/dashboard: Use ModalComponent in all modals

Fixes: https://tracker.ceph.com/issues/44547
Signed-off-by: Tiago Melo <tmelo@suse.com>

Merge pull request #33902 from dillaman/wip-test-rbd-integration

test/cli-integration: fixed spacing issue for RBD formatted tables

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

mgr/cephadm: reconfigure ceph daemons when monmap changes

This keeps ceph.conf fresh.

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

Merge pull request #33898 from dillaman/wip-44567

qa/workunits/rbd: wait for nbd map to close after unmap

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

cephadm: add version to `command_ls` output

Signed-off-by: Michael Fritch <mfritch@suse.com>

cephadm: add type checking to `update_filewalld`

fixes mypy errors:
cephadm:1682: error: Incompatible types in assignment (expression has type "str", variable has type "int")
cephadm:1683: error: List item 3 has incompatible type "int"; expected "str"
cephadm:1686: error: List item 3 has incompatible type "int"; expected "str"
Found 3 errors in 1 file (checked 1 source file)

Signed-off-by: Michael Fritch <mfritch@suse.com>

cephadm: allow prepare-host to start an enabled service

the service might be enabled, but not running

Signed-off-by: Michael Fritch <mfritch@suse.com>

cephadm: add type checking for `check_host` and `prepare_host`

Signed-off-by: Michael Fritch <mfritch@suse.com>

cephadm: generalize logic for checking and enabling units

Signed-off-by: Michael Fritch <mfritch@suse.com>

cephadm: add 'CEPH_CONF' to the NFS ganesha container envs

Signed-off-by: Michael Fritch <mfritch@suse.com>

cephadm: trim nfs.json sample

we only need to specify the rados pool and config object url

Signed-off-by: Michael Fritch <mfritch@suse.com>

qa/workunits/cephadm/test_cephadm.sh: systemctl stop nfs-server

stop any running nfs servers before attempting a deploy of nfs-ganesha

Signed-off-by: Michael Fritch <mfritch@suse.com>

qa/workunits/cephadm/test_cephadm.sh: make pgs available

allow rados obj. create to succeed during a single node test

Signed-off-by: Michael Fritch <mfritch@suse.com>

cephadm: add some log lines

Signed-off-by: Michael Fritch <mfritch@suse.com>

cephadm: check port in use

Signed-off-by: Michael Fritch <mfritch@suse.com>

cephadm: add/remove nfs ganesha grace

container unit pre-start/post-stop for ganesha grace db

Signed-off-by: Michael Fritch <mfritch@suse.com>

cephadm: update firewalld with nfs service

Signed-off-by: Michael Fritch <mfritch@suse.com>

qa/workunits/cephadm/test_cephadm.sh: add nfs-ganesha test

Signed-off-by: Michael Fritch <mfritch@suse.com>

cephadm: add ganasha.conf

also allows for any arbitrary nfs conf to be deployed by adding items to
the `files` dict of the config-json

Signed-off-by: Michael Fritch <mfritch@suse.com>

cephadm: add NFSGanesha deployment type

Signed-off-by: Michael Fritch <mfritch@suse.com>

cephadm: consolidate list of supported daemons

Signed-off-by: Michael Fritch <mfritch@suse.com>

cephadm: use keyword instead of positional args

Signed-off-by: Michael Fritch <mfritch@suse.com>

mgr/rbd_support: fix recursive non-global level schedule listing

For the case when the non-global level does not have a schedule
and a higher level is used as the parent, it wrongly listed
schedules from all branches under the parent, instead of only the
interested one.

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

mgr/dashboard: Updated octopus image on 404 page

Replaced the image of the Nautilus octopus with another octopus
in preparation for the "Octopus" release.

The image was taken from Museums Victoria
(https://collections.museumvictoria.com.au/species/8696) and is
licensed under the Creative Commons "Attribution 4.0 International"
(CC BY 4.0) license.

Deleted older, now obsolete images from the assets directory.

Fixes: https://tracker.ceph.com/issues/43581
Signed-off-by: Lenz Grimmer <lgrimmer@suse.com>

doc: Improved Dashboard installation workflow

Moved the "Accessing the Dashboard" section in front of
the paragraphs that talk about enabling additional services.

Signed-off-by: Lenz Grimmer <lgrimmer@suse.com>

doc: Updated dashboard feature overview for Octopus

Added new Dashboard features added for the Octopus
release.

Signed-off-by: Lenz Grimmer <lgrimmer@suse.com>

mgr/orch: Cleanup python imports of the interface

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>

mgr/orch: Properly handle NotImplementedError

They are masked as RuntimeError by the mgr.

Fixes: https://tracker.ceph.com/issues/44569
Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>

pybind/mgr: Rm PersistentStoreDict

`PersistentStoreDict` works great, except that is
too expensive to be used.

In addition, no one makes use of it.

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>

cephadm: log when we infer the image

This is similar to how we log when we infer the fsid.

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

cephadm: Avoid `inspect` call by using ceph label

Signed-off-by: Ricardo Marques <rimarques@suse.com>

cephadm: Infer ceph image

Fixes: https://tracker.ceph.com/issues/44440
Signed-off-by: Ricardo Marques <rimarques@suse.com>

Merge PR #33897 into octopus

* refs/pull/33897/head:
mgr/PyModule: initialize options on standby class too
Revert "pybind/mgr/mgr_module: fix standby module logging options"

Reviewed-by: Ernesto Puerta <epuertat@redhat.com>
Reviewed-by: Sebastian Wagner <swagner@suse.com>

osd/PGStateUtils: initialize NamedState::enter_time.

This bug introduce by d96c907.

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

osd/PGStatueUtils: remove unused member variable

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

qa/workunits/rbd: wait for schedule status updated

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

Merge PR #33870 into octopus

* refs/pull/33870/head:
mgr/cephadm: fix redeploy for the crash daemon
cephadm: prefer config/keyring args
cephadm: raise Error when no config/keyring
cephadm: consolidate logic for reading `config-json`
cephadm: merge `--config-and-keyring` and `--config-json` args

Reviewed-by: Sage Weil <sage@redhat.com>
Reviewed-by: Sebastian Wagner <swagner@suse.com>

mgr/dashboard: Create bucket with x-amz-bucket-object-lock-enabled

Fixes: https://tracker.ceph.com/issues/43446
Signed-off-by: Volker Theile <vtheile@suse.com>

Merge pull request #33831 from dillaman/wip-44396

rbd-mirror: prevent I/O modifications against a non-primary image

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

Merge pull request #32419 from rhcs-dashboard/fix-43419-master

mgr/dashboard: add feature toggle for NFS and fix feature toggles regression

Reviewed-by: Alfonso Martínez <almartin@redhat.com>
Reviewed-by: Laura Paduano <lpaduano@suse.com>
Reviewed-by: Stephan Müller <smueller@suse.com>
Reviewed-by: Volker Theile <vtheile@suse.com>

mgr/PyModule: initialize options on standby class too

Create a separate callback (_register_options) and call that on both
the regular MgrModule and MgrStandbyModule.

Fixes: https://tracker.ceph.com/issues/44562
Signed-off-by: Sage Weil <sage@redhat.com>

Merge PR #33817 into octopus

* refs/pull/33817/head:
mgr/dashboard: Adapt tests to new DriveGroupSpec
fixup mgr/test_orchestrator: validate drive group matches anything.
mgr/orch: CLI: No Tracebacks for ServiceSpecValidationError
mgr/test_orchestrator: validate drive group matches anything.
python-common: don't run flake8 on tests.
python-common: Add support for legacy serialization format for Drive Groups
doc: Move Move ServiceSpec to python-common
python-common: Add `host_pattern` to `PlacementSpec.from_string()`
cephadm: add host_pattern to supported scheduling
python-common: Joined ServiceSpec and DriveGroupSpec from_json()
python-common: Make DriveGroupSpec a sub type of ServiceSpec
pybind/mgr: Move ServiceSpec to python-common: Fix imports
python-common, orch: Move ServiceSpec to python-common: Fix imports
python-common, orch: Move ServiceSpec tests to python-common
python-common: Move ServiceSpec to python-common: fix linting
python-common, orch: Move ServiceSpec (+deps) to python-common

Reviewed-by: Joshua Schmid <jschmid@suse.de>

Merge PR #33895 into octopus

* refs/pull/33895/head:
cephadm: update unit.* atomically

Reviewed-by: Michael Fritch <mfritch@suse.com>

cephadm: update unit.* atomically

Some of these are run as bash scripts, which means that updating them
can lead to the running bash picking up at a weird position mid-script
when it goes to the next command.  This produces weird errors like

bash[9321]: /var/lib/ceph/f1758250-639e-11ea-9a42-001a4aab830c/mon.c/unit.run: line 2: -to-stderr=true: command not found

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

test/cli-integration: fixed spacing issue for RBD formatted tables

The tables now use two spaces between columns instead of one.

Signed-off-by: Jason Dillaman <dillaman@redhat.com>

mgr/cephadm: fix redeploy for the crash daemon

`auth get` was using `crash.<name>` when it should have been
`client.crash.<name>`

Signed-off-by: Michael Fritch <mfritch@suse.com>

cephadm: prefer config/keyring args

let `--config` or `--keyring` take precedence over the values supplied
via the `--config-json`

Signed-off-by: Michael Fritch <mfritch@suse.com>

cephadm: raise Error when no config/keyring

fixes mypy error:
cephadm:1141: error: Incompatible return value type (got "Tuple[Optional[str], Optional[str]]", expected "Tuple[str, str]")

Signed-off-by: Michael Fritch <mfritch@suse.com>

cephadm: consolidate logic for reading `config-json`

logic for reading `--config-json` from stdin or file was duplicated

Signed-off-by: Michael Fritch <mfritch@suse.com>

cephadm: merge `--config-and-keyring` and `--config-json` args

both are passed via stdin from the mgr in a JSON format

Signed-off-by: Michael Fritch <mfritch@suse.com>

qa/suites/rbd: removed OpenStack tempest test cases

The OpenStack tempests tests do not stay stable and break approximately
every six months. Remove the test suite for now.

Signed-off-by: Jason Dillaman <dillaman@redhat.com>

qa/workunits/rbd: wait for nbd map to close after unmap

The unmap action only sends a signal to the kernel to notify the
rbd-nbd daemon to disconnect. Therefore, it's possible that an
unmap followed by an immediate re-map to the same device might
fail since the unmap is still in-progress.

Fixes: https://tracker.ceph.com/issues/44567
Signed-off-by: Jason Dillaman <dillaman@redhat.com>

Merge pull request #33848 from mchangir/octopus-tests-remove-suprious-whitespace

tests: remove spurious whitespace

Revert "pybind/mgr/mgr_module: fix standby module logging options"

This reverts commit abcee7133b87d60c841ad2ed589858e4dd347ed8.

Breaks the reporting of these options up through the mgr daemon,
because PyModule::load_options() happens before the module ctor
excutes.

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

doc/install: rethink docs install page (similar to PR#33613)

This rearranges the list of installation methods, in accordance with Sage's and Sebastian W's wishes.

Signed-off-by: Zac Dover <zac.dover@gmail.com>

Merge PR #33746 into octopus

* refs/pull/33746/head:
cephadm: add grafana adopt

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

mgr/dashboard: add feature toggle for NFS

- New NFS feature toggle added.
- Fixed regression which broke FeatureToggles in the main menu.
- Added extensive unit testing to NavigationComponent
- Added ng-mocks to improve test isolation and performance (~7x, 139s to
20s)
- Added ng-bullet package to improve unit testing performance (2x, 20s
to 9s)
- Added Rxjs Schedulers to implement NgZone.runOutsideAngular in a Rxjs
friendly way (based on
    https://stackoverflow.com/questions/43121400/run-ngrx-effect-outside-of-angulars-zone-to-prevent-timeout-in-protractor)

Minor issues found from exhaustive unit testing:
- Missing permissions in Cluster menu:
  - `permissions.log.read` and `permissions.prometheus.read`
- Missing classes:
  - Block -> Images: `tc_submenuitem_block_images`
  - Block -> iSCSI: `tc_submenuitem_block_iscsi`
- Typos:
  - class `tc_submenuitem_hosts` assigned to OSD menu (instead of
      `tc_submenuitem_osd`)
  - `tc_menuitem_cephs` -> `tc_menuitem_cephfs`

Minor changes:
- Previously, Cluster Map -> CRUSH Map required both OSD and Host
permissions. Now it only requires OSD permissions (there are no
references to hosts in CRUSH Map page). Nevertheless, all system roles
setting OSD permission also set Host's, so no impact expected.
- Previously, Cluster -> Monitoring menu was hidden when both Prometheus
or Alertmanager weren't configured. Now it's displayed and when clicked
on it shows the helper banner indicating that further configuration is
required. This change removes the dependency on the PrometheusService.

Fixes: https://tracker.ceph.com/issues/43419
Fixes: https://tracker.ceph.com/issues/43715
Signed-off-by: Ernesto Puerta <epuertat@redhat.com>

Merge PR #33830 into octopus

* refs/pull/33830/head:
qa/tasks/cephadm: no default mon|mgr|crash service specs
qa/suites/rados/cephadm/upgrade: upgrade start point that supports the no-spec option
cephadm: create initial mon and mgr service specs too
cephadm: no need to pregenerate a crash key for the bootstrap host
mgr/cephadm: do not complain when we don't have enough hosts
mgr/cephadm: remove orphan daemons
mgr/cephadm: report size=0 for fabricated ServiceDescription
mgr/cephadm: safety check to prevent removing all mon|mgr daemons
mgr/cephadm: prevent scaling mon|mgr below count=1
mgr/cephadm: do not remove daemons from remove_service

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

Merge PR #33732 into octopus

* refs/pull/33732/head:
mon/MonClient: send logs to mon on separate schedule than pings
mon/MonClient: send logs to mon even if we have no keelalive2

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

Merge PR #33620 into octopus

* refs/pull/33620/head:
mgr/dashboard: Crush rule modal
mgr/dashboard: Preserve rule selection on pool type change
mgr/dashboard: Crush rule is only send during replicated pool creation
mgr/dashboard: Explicit returns in pool form
mgr/dashboard: Removes fork join in pool form
mgr/dashboard: Hide ECP actions during ec pool edit
mgr/dashboard: Pool form erasure/replicated boolean
mgr/dashboard: Change pool info API endpoint
mgr/dashboard: Moves ECP info endpoint to UI-API

Reviewed-by: Tiago Melo <tmelo@suse.com>

test/rbd: fix test_rbd.test_features_to_string

The new RBD_FEATURE_NON_PRIMARY overlaps with the legacy test
for an invalid feature bit.

Signed-off-by: Jason Dillaman <dillaman@redhat.com>

qa/suites/rados/verify: debug_ms = 1, osd_heartbeat_grace = 60

The rados api tests are failing WatchNotify because the OSDs are so
heavily lagged.. in large part due to the high debug level of debug_ms=20
and debug_osd=25.  Reduce that.

Also increase the heartbeat grace so slow valgrind-y osds don't get marked
down.

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

qa/suites/powercycle: install build deps for building xfstest

see also qa/suites/krbd/rbd/tasks/rbd_workunit_suites_fsx.yaml

Fixes: https://tracker.ceph.com/issues/44552
Signed-off-by: Kefu Chai <kchai@redhat.com>