mgr/dashboard: Fix doc endpoints

UiApi endpoints were broken because the same base URL was used for
both UiApi and Api endpoints.

There were two issues with this:

- The paths dict keys were based off a string split of path that
  used base URL length. This caused the wrong keys to be created and the
duplicate apiapi seen in the cURL requests generated by OpenAPI.
- Fixing the above issue was not enough, as then both Api and UiApi
  endpoints would have the same base URL. This causes conflicts with doc
generation that result in only UiApi endpoints to be generated.

Passing only / as the base URL argument for api_json and
api_all_json and using the full path as the paths keys solves this
issue.

Other changes:

* base_url variable was removed from  controllers/docs.py:_gen_paths
  and tests/test_docs.py because it was unused after my change.

* Added type hints for dicts in controllers/docs.py and
  controllers/pool.py because they were not passing mypy checks.

* test_docs.py unit test added to verify change

* Add message which explains that UiApi endpoints are not part of the
public api

Fixes: https://tracker.ceph.com/issues/45957
Signed-off-by: Fabrizio D'Angelo <fdangelo@redhat.com>
(cherry picked from commit d1a5300a6e778d8fca5befbc2023df0f5c3ead03)

Merge pull request #36046 from smithfarm/wip-46469-octopus

octopus: cephfs: client: release the client_lock before copying data in read

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

Merge pull request #36045 from smithfarm/wip-46410-octopus

octopus: cephfs: client: fix setxattr for 0 size value (NULL value)

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

Merge pull request #36044 from smithfarm/wip-46348-octopus

octopus: tests: qa/tasks: make sh() in vstart_runner.py identical with teuthology.orchestra.remote.sh

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

Merge pull request #36043 from smithfarm/wip-46311-octopus

octopus: tests: qa/tasks/cephfs/cephfs_test_case.py: skip cleaning the core dumps when in program case

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

Merge pull request #36042 from smithfarm/wip-46199-octopus

octopus: tests: Revert "Revert "qa/suites/rados/mgr/tasks/module_selftest: whitelist …

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

Merge pull request #36040 from smithfarm/wip-46188-octopus

octopus: mds: reset heartbeat in EMetaBlob replay

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

Merge pull request #36039 from smithfarm/wip-46186-octopus

octopus: cephfs: client: fix snap directory atime

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

Merge pull request #36038 from smithfarm/wip-46152-octopus

octopus: mds: do not defer incoming mgrmap when mds is laggy

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

Merge pull request #36035 from smithfarm/wip-46286-octopus

octopus: mon/MDSMonitor: copy MDS info which may be removed

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

Merge pull request #35842 from Vicente-Cheng/wip-46190-octopus

octopus: mds: fix filelock state when Fc is issued

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

Merge pull request #36109 from sebastian-philipp/octopus-backport-35890-35913-35908-35927-35813-35717-35990-35915-35747-36013

octopus: cephadm batch backport July (2)

Reviewed-by: Adam King <adking@redhat.com>
Reviewed-by: Juan Miguel Olmo Martínez <jolmomar@redhat.com>
Reviewed-by: Kiefer Chang <kiefer.chang@suse.com>

Merge pull request #35819 from bk201/wip-46197-octopus

octopus: mgr/dashboard: fix Source column i18n issue in RBD configuration tables

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

Merge pull request #36073 from rhcs-dashboard/wip-46354-octopus

octopus: mgr/dashboard: fix backporting issue #35926

Reviewed-by: Alfonso Martínez <almartin@redhat.com>
Reviewed-by: Kiefer Chang <kiefer.chang@suse.com>
Reviewed-by: Laura Paduano <lpaduano@suse.com>

Merge pull request #35997 from ricardoasmarques/wip-46436-octopus

octopus: mgr/dashboard: Allow to edit iSCSI target with active session

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

Merge pull request #35684 from smithfarm/wip-46121-octopus

octopus: mgr/k8sevents: sanitise kubernetes events

Merge pull request #35958 from callithea/wip-46394-octopus

octopus: mon/PGMap: do not consider changing pg stuck

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

cephadm: Make Vagrantfile more flexible

Now you can use a JSON or pass multiple variable to vagrant in order to
configure the outcome of VMs you get. Similar to vstart.sh you can use
OSDS, MGRS and MONS as arguments to pass. As OSDS behave a bit different
in this scenario you can also specify the amount of extra disks and OSD
VM has.

Fixes: https://tracker.ceph.com/issues/46376
Signed-off-by: Stephan Müller <smueller@suse.com>
(cherry picked from commit c767a0c0e8ffed4448e4d2cacef72674e7ada883)

mgr/cephadm: allow custom dashboard grafana url

Instead of resetting grafana url to a default with
each run of the service loop, only set it when
initally deployed or when it's placement changes

Fixes: https://tracker.ceph.com/issues/44877
Signed-off-by: Adam King <adking@redhat.com>
(cherry picked from commit fc14c11b4f767b46794ab6b5eb70d53ea38af384)

mgr/rook: Fix error creating OSDs

https://github.com/ceph/ceph/pull/34860 broke Rook Integration tests (Rook orchestrator module)

This fix the error that can be seen in:
https://jenkins.rook.io/blue/rest/organizations/jenkins/pipelines/rook/pipelines/rook/branches/master/runs/2046/nodes/63/steps/121/log/?start=0

Signed-off-by: Juan Miguel Olmo Martínez <jolmomar@redhat.com>
(cherry picked from commit 9fd4b481fc4d18e9794806167588d6c6cb1aa46a)

cephadm: Fix Python formating issues

Fix obvious Python formating errors reported by my IDE.

Signed-off-by: Volker Theile <vtheile@suse.com>
(cherry picked from commit 906fa6da4a602d097b740a4d0146bef2d8dd5c59)

Conflicts:
src/cephadm/cephadm

doc/dev/cephadm.rst: clarify, fix spelling nits

Signed-off-by: Dan Mick <dmick@redhat.com>
(cherry picked from commit d108444fedeffd9f2268e5277af58f2b1030aa33)

vstart.sh: use output of hostname for cephadm

Otherwise I get this error on a dev machine:

        /home/pdonnell/ceph/build/bin/ceph -c /home/pdonnell/ceph/build/ceph.conf -k /home/pdonnell/ceph/build/keyring orch host add senta03
        Error ENOENT: New host senta03 (senta03) failed check: ['INFO:cephadm:podman|docker (/bin/podman) is present', 'INFO:cephadm:systemctl is present', 'INFO:cephadm:lvcreate is present', 'INFO:cephadm:Unit chronyd.service is enabled and running', 'INFO:cephadm:Hostname "senta03" matches what is expected.', 'ERROR: hostname "senta03.front.sepia.ceph.com" does not match expected hostname "senta03"']

If `hostname` is configured to give the fqdn, we get the above error
from cephadm.

Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
(cherry picked from commit 4aba413337e4f9041ade9b5b9e9daee9a08eb419)

qa/tasks: fix cephadm config image check

Check container_image_name only if ceph cluster image is not pre-defined in config.
We shouldn't care about container_image_name if there cephadm or ceph already have image defined.

Signed-off-by: Georgios Kyratsas <gkyratsas@suse.com>
(cherry picked from commit 390ca4d92fe2e0467eec9161de1b7bdc1c62891f)

mgr/cephadm: fix Dashboard ganesha setting duplication error

Fix the problem when there are multiple daemons in a NFS service,
redundant Ganesha locations are set.

This change also fixes the setting is wrong when no namespace is used for
a Ganesha config pool.

Fixes: https://tracker.ceph.com/issues/46329
Signed-off-by: Kiefer Chang <kiefer.chang@suse.com>
(cherry picked from commit c233ef9099685276de39e799469aebfcb438b1ea)

docker/podman: tag crimson flavor container as 'sha-1-flavor'

see also: ceph/ceph-container#1704

Signed-off-by: Deepika Upadhyay <dupadhya@redhat.com>
(cherry picked from commit b0a14d6e1eb194fe654fe160430d953a96a7e355)

cephadm: Make list_networks ipv6 enabled

Currently the list_network command and methods in cephadm only run and
parse ipv4 output from `ip route`.

This patch extends the list_network command and internal methods to be
ipv6 enabled. It now also checks `ip -6 route` and `ip -6 addr` to
gather gather all networks from both protocol families.

Signed-off-by: Matthew Oliver <moliver@suse.com>
(cherry picked from commit c64d273084bcd3d43c6b63dc070de3f244f86ca8)

Merge pull request #36034 from smithfarm/wip-46261-octopus

octopus: common, osd: add sanity checks around osd_scrub_max_preemptions

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

Merge pull request #36033 from smithfarm/wip-46089-octopus

octopus: osd/PG: fix history.same_interval_since of merge target again

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

Merge pull request #36032 from smithfarm/wip-46086-octopus

octopus: osd: wakeup all threads of shard rather than one thread.

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

Merge pull request #36030 from smithfarm/wip-46016-octopus

octopus: test: osd-backfill-stats.sh use nobackfill to avoid races in remainin…

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

Merge pull request #36029 from smithfarm/wip-46007-octopus

octopus: osd/PrimaryLogPG: don't populate watchers if replica

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

Merge pull request #35898 from sebastian-philipp/octopus-backport-35018-35472-35524-35412-35510-35493-35549-35547-35581-35567-35578-3563

octopus: cephadm batch backport July (1)

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

cephadm: Add "--format" option to "ceph orch status"

Fixes: https://tracker.ceph.com/issues/46233
Signed-off-by: Ricardo Marques <rimarques@suse.com>
(cherry picked from commit 83ff83157c3c5bc3c825d09c8724310e02aea988)

mgr/test_orchestrator: add iscsi

Signed-off-by: Michael Fritch <mfritch@suse.com>
(cherry picked from commit 37826d41b0bcf755fc9a837c72741542022a019a)

cephadm: add sudo ssh user option

add the ability to use a non-root users for connecting to hosts, this user requires pass wordless sudo access

useful for clusters that have root ssh access disabled

Fixes: https://tracker.ceph.com/issues/44866
Signed-off-by: Daniel-Pivonka <dpivonka@redhat.com>
(cherry picked from commit 1052b9db7bdc50f61705e1f1331c5cbb20120586)

mgr/cephadm: Manage /etc/ceph/ceph.conf

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit c18ad7c8a3e63e6eb26947bb7ab47078ebfd0d60)

mgr/tests: _ceph_get_module_option returns proper type now.

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit acdd26ab4e30ff5abee1d555c5e45787e3d4f02f)

mgr/cephadm: config_notify now provides a hook

reason is, we want to use this hook to schedule a
ceph.conf update for all hosts.

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit 1cd4d0f32380e7ad0fbaf365df5f45a343fcb458)

mgr/cephadm: refactor _run_cephadm

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit a0c39f3dda40e2ce6e3877fcd1176f5bb003a542)

python-common: Add simple str(HostSpec)

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit e60a2a5dede1ad4e4f4c9e9aefc402032d28320c)

mgr/cephadm: add type hints for _refresh_host_*()

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit c37412deb0598a5985da8097d6891d4516601c10)

cephadm: ceph-iscsi remove pool from cap

When we create a ceph-iscsi daemon/continer in cephadm we create a user
and set some caps. Turns out we were a little too restrictive.

We were locking down to only access the pool that was given in the spec,
which happens to be the pool the iscsi config is stored. But in reality
we need to be able  to attach any rbd images which could exist in other
pools.

So this patch removes the `pool=` from the osd cap, so from:

   osd = allow rwx pool={spec.pool}

To:

   osd = allow rwx

Fixes: https://tracker.ceph.com/issues/46138
Signed-off-by: Matthew Oliver <moliver@suse.com>
(cherry picked from commit 8cf51251a3299bf5a65ea338f9fb06c4f3052ad1)

py-common/deployment/translate: drop unneeded ctor arg

The DriveGroupSpec is already part of the DriveSelection.

Fixes: https://tracker.ceph.com/issues/46231
Signed-off-by: Jan Fajerski <jfajerski@suse.com>
(cherry picked from commit 1a485406da49977a1fc83dda8204b9b94634bec1)

mgr/cephadm: reconfig ssh after setting or clearing ssh_config

Fixes: https://tracker.ceph.com/issues/46245
Signed-off-by: Kiefer Chang <kiefer.chang@suse.com>
(cherry picked from commit 027b4e154a70b82408de1b19d34a7d4706acded7)

python-common: service_spec.from_json: Add docstring explaining histroy

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit 5150e957844354861f8adc916f87b6b15ccb3d6a)

qa/cephadm: Add test for --format=yaml

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit 1650257015b9470918f50478a6151f2f40467dd3)

mgr/orchestrator: Add test for readable yaml

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit 87a150358eb0a1cfa8b38a6df4135b57a21f7130)

pybind/mgr: move test_orchestrator to orchestrator module

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit 888addc4680ea8fb6c9f2c88c418c53e910a2ba9)

mgr/orch: make YAML of ServiceDescription readable

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit fd29e1c4a8ccaa84f3f2e73920b637576d565015)

mgr/orch: YAML of DaemonDescription readable

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit 4c7ad07b20fd5ae125bad34fcebdd84a886432b9)

python-common: Make YAML or OSD Specs readable

* Changes: An empty OSD Spec is now invalid.
* OSDSpec.validate() now fails, if service-id is empty

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit cd6a488ab2ca036dd4fb36751b938f605e97e1c8)

python-common: make YAML representaition of ServiceSpec readable

* Add test for new yaml representation

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit 4ddb111dd1491a4456bdc0379b6fbc5d641f4291)

cephadm: error trying to get ceph auth entry for crash daemon

If your cluster has nodes with a . in the name. This will happen.

Signed-off-by: Juan Miguel Olmo Martínez <jolmomar@redhat.com>
(cherry picked from commit 68cfa109b46bb627cfdfe624815aa01ed6b6c39b)

mgr/cephadm: add test for ok-to-stop

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit 5c91f1b897beaca476bb1ef178ef332fe092bae3)

mgr/cephadm: scheduler: make use of ok-to-stop

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit f42182587057e8e18661f0b7570c7af1617e3506)

mgr/cephadm: move _check_safe_to_destroy_mon to MonService

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit cb92429d52319deb2bc626495ea58ac82e50e558)

mgr/cephadm: move ok_to_stop to CephadmService

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit 2b41d0a470c91d4395a2af4ee88b2f26253791ca)

mgr/cephadm: use fixed Grafana version

Fixes: https://tracker.ceph.com/issues/45859
Signed-off-by: Patrick Seidensal <pseidensal@suse.com>
(cherry picked from commit 571960562187d8588826c6616e90e23948376e8d)

mgr/cephadm: speedup _refresh_hosts_and_daemons by 10x

by calling 10 hosts in parallel.

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit 0716b0556996ce34b628611274064cbf05fef97c)

mgr/cephadm: fix broken "already in use" exception message

Fixes:

```
orchestrator._interface.OrchestratorValidationError: ('name %s already in use', 'b')
```

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit 9a45c39c68956c87e55ec4cce90a7bb4df524b97)

cephadm: Set ms bind ipv6 when mon-ip is ipv6

If you use cephadm bootstrap with an ipv6 mon ip then currently you'll
get into a address family split-brain state, where the mon's messenger
connects and binds to ipv6 but the mgr's binds to ipv4 (usually
0.0.0.0). In this state the bootstrap process hangs as it attempts to
talk and get the mgr state.

A work around is to have `ms bind ipv6 = true` in a ceph conf
you can then pass to bootstrap which gets pulled in and set in mon's
config store.

This patch sets `ms bind ipv6 = true` to the global section in the
mon config store when the mon-ip argument is an ipv6 address.

Fixes: https://tracker.ceph.com/issues/45016
Signed-off-by: Matthew Oliver <moliver@suse.com>
(cherry picked from commit 08ba08f7bb5b577ad3c3895e2c7f9f4d4555f185)

mgr/cephadm: exclude osds without osdspec_affinity

Signed-off-by: Joshua Schmid <jschmid@suse.de>
(cherry picked from commit 343cf0386264e0c6c4851614b7251c18792e2d8b)

cephadm: restrict mds caps to cephfs pools

Fixes: https://tracker.ceph.com/issues/46081
Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
(cherry picked from commit 1ef958085d1fa86f3c79ba25b0e20be2e6fc5dd5)

cephadm: unit.run: add `set -e`

In case LVM activates fails for some
unknown reason, prevent the script from
continuing with starting the OSD container.

This leads to faild systemd services with active OSD containers.

Fixes: https://tracker.ceph.com/issues/46036
Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit f9e5dd509bcc5d5b7c9849f58183a9de6f56ee1e)

mgr/orch: read untrusted input using the yaml SafeLoader

https://msg.pyyaml.org/load

Signed-off-by: Michael Fritch <mfritch@suse.com>
(cherry picked from commit d5175e3a443a4e17b6f73a2ced35d224ab4d93e2)

mgr/cephadm: make `orch status` fail, if ssh key is not set

ceph-salt relies on the output of `orch status` to determine the
end of the bootstrap process.

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit f4e7d33d173f6e1236af531fac8e7a2a2c27af53)

python-common: remove old `cephadm_orchestrator.py`

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit 0111c880f4951ed0ee395190a9bf67bd7ebb0228)

doc/cephadm: adopt: fix misleading `apply` examples

Fixes: https://tracker.ceph.com/issues/46052
Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit e74fc0671b73b6da3669f5389f372342d4e62843)

doc/mgr/orch: Update documentation to the new CLI

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit b0bbef91f0fb0afcd962f1ea26e99faa5031cf52)

cephadm: Give iscsci a RO /lib/modules bind mounted

The ceph iscsi container needs to be able to insert the iscsi_target_mod
but it doesn't exist in the container. for security reasons bind
mounting /lib/modules seems a little dangerous unless we can mount it
RO.
Unfortuntly the docker volume mount (-v) doesn't allow you mount
readonly, adding a `--read-only` actaully does the opposite, makes the
root on the container RO and expects you to write to the mounted volumes
(-v).

However, we get more grainular control over bind mount options if we use
`--mount`[0]. Here we can still bind mound the volume into the container,
but can also add additional options, like bind mounting RO.

This patch adds at addiontal `bind_mounts` option to the CephContainer
along side `volume_mounts`. The `bind_mounts` take a List[List[str]]:

 binds = []
 lib_modules = ['type=bind',
                'source=/lib/modules',
                'destination=/lib/modules',
                'ro=true']
 binds.append(lib_modules)

And this is plumbed through into cephadm. Bind_mounts only needs to be
used if you need a little more control over the mounting, otherwise the
volume_mounts are easier to use.

[0] - https://docs.docker.com/engine/reference/commandline/service_create/#add-bind-mounts-volumes-or-memory-filesystems

Fixes: https://tracker.ceph.com/issues/45252
Signed-off-by: Matthew Oliver <moliver@suse.com>
(cherry picked from commit d9b5371478b744920cf14e1b34b7d63226c71050)

stop.sh: do not block script when there is no running cluster

A query for current fsid is called inside `do_killcephadm`. This blocks
the script when there is no running cluster. The fix avoids entering the
function if cephadm command fails or returns no daemons.

The change also hides the following output for non-cephadm environments:

```
Unable to locate any of ['podman', 'docker']
```

Fixes: https://tracker.ceph.com/issues/46067
Signed-off-by: Kiefer Chang <kiefer.chang@suse.com>
(cherry picked from commit 9884b9313db506a35b4e49d46c754965d90476ed)

doc/man: add missing cephadm shell `--mount` option

also clean-up `--mount` help usage

Signed-off-by: Michael Fritch <mfritch@suse.com>
(cherry picked from commit 00baed29f52965e17bb0a7d9db7f9a5b0f696290)

cephadm: skip any invalid fsid(s) during infer

Fixes: https://tracker.ceph.com/issues/45999
Signed-off-by: Michael Fritch <mfritch@suse.com>
(cherry picked from commit d2e8b812051323cdea0d594df1d2ae2cb3e287b1)

cephadm: sort the list of inferred fsids

$ cephadm shell
ERROR: Cannot infer an fsid, one must be specified: ['1d5df33f-eb94-4a4f-b192-1d5e770ed0e7', 'unknown']
$ cephadm shell
ERROR: Cannot infer an fsid, one must be specified: ['unknown', '1d5df33f-eb94-4a4f-b192-1d5e770ed0e7']

Signed-off-by: Michael Fritch <mfritch@suse.com>
(cherry picked from commit 320956eb5ea5a43f0a0cd93fe580bc8d699eff1b)

mgr/cephadm: annotate scheduling code with comments

Signed-off-by: Joshua Schmid <jschmid@suse.de>
(cherry picked from commit 856ab3bdcf9a4ee364377c569490cca8094a4b04)

doc/cephadm: document pool requirement for NFS-Ganesha service

Signed-off-by: Kiefer Chang <kiefer.chang@suse.com>
(cherry picked from commit 71140f47e50d4aa95aa56996e9bafbdf6ad65b14)

cephadm: UX: Change error message when 'orch host add <host>' fails

Instead of printing out a traceback if adding the host fails
during bootstrapping process, should now print error message
telling user host failed to be added

Fixes: https://tracker.ceph.com/issues/45097
Signed-off-by: Adam King <adking@redhat.com>
(cherry picked from commit 0e2dd36aec24f54d51af5551a75a184fdefd325d)

cephadm: strip whitespace from iscsi version

Signed-off-by: Michael Fritch <mfritch@suse.com>
(cherry picked from commit f23649cb5412735e37f3775ebb086f1dc6c7d235)

python-common: validate the service_id during apply

the service_id needs to be validated during `orch apply`, but not during
`orch daemon add`

Signed-off-by: Michael Fritch <mfritch@suse.com>
(cherry picked from commit c83144517caaf217f7f10573842f12472bef1cc2)

python-common: Add missing IscsiServiceSpec test

Signed-off-by: Michael Fritch <mfritch@suse.com>
(cherry picked from commit dcf7fb520e3e55fbb5daf035c0cb7525e899b5b0)

python-common: Add NFSServiceSpec validate

the ServiceSpec needs to be validated during `orch apply`, but not
during `orch daemon add`

Signed-off-by: Michael Fritch <mfritch@suse.com>
(cherry picked from commit 45391a1270178de87c6d2305f657809d5e1e6072)

cephadm: only SIGKILL once during timeout

no need to continually SIGKILL a process that is already exiting

Signed-off-by: Michael Fritch <mfritch@suse.com>
(cherry picked from commit 1b637144a6bbbeb752f1a006ce0f7647b084783a)

cephadm: add profile log line to Popen wrappers

add debug log line to profile the Popen wrapper read loop
exec runtime, exitcode, stop, fds to read, etc.

Signed-off-by: Michael Fritch <mfritch@suse.com>
(cherry picked from commit 49887ab4554576249fe6c205a46bff2a674b5ec2)

cephadm: increase is_available timeout 30s -> 60s

bootstrap fails because `ceph -s` might take longer
than 30 sec to return on resource limited hardware

Fixes: https://tracker.ceph.com/issues/45961
Signed-off-by: Michael Fritch <mfritch@suse.com>
(cherry picked from commit de734aa05fdad770c96c70d4b1b6cdddc98ce71b)

doc/cephadm: Add chapter about FQDNs

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit 55e608d796952881083f3620c91edb5596da8220)

cephadm: call `podman rm --storage`

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit decfbca835ae0a2f03ae169f79fafa1d0babbbcb)

stop.sh: Remove cephadm deployed daemons

Fixes: https://tracker.ceph.com/issues/44746
Signed-off-by: Varsha Rao <varao@redhat.com>
(cherry picked from commit 4930dc71b5d808c987a5d5f1d8ed62c8f0c87765)

cephadm: Manually remove containers

This fixes:
```
Error: error creating container storage: the container name "ceph-<fsid>-mon.b" is already in use by "<container-id>". You have to remove that container to be able to reuse that name.: that name is already in use
```

Relates to https://tracker.ceph.com/issues/44990

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit b3fbc0329c15d0a7d73d9e0ad6f0065d724b12a7)

Merge pull request #35988 from votdev/adapt_cephadm_vagrant

octopus: mgr/cephadm: Adapt Vagrantfile to use octopus instead of master repo on shaman

Reviewed-by: Stephan Müller <smueller@suse.com>

Merge pull request #36063 from varshar16/wip-update-nfs-doc-backport

octopus: Update nfs-ganesha package requirements doc backport

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

mgr/dashboard: fix backporting issue #35926

Incorrect conflict-solving during backporting of https://github.com/ceph/ceph/pull/34606
to Octopus (https://github.com/ceph/ceph/pull/35926) led to deletion of
some chuncks in OSD list page.

Fixes: https://tracker.ceph.com/issues/46502
Fixes: https://github.com/ceph/ceph/pull/35926
Signed-off-by: Ernesto Puerta <epuertat@redhat.com>

Merge pull request #36061 from idryomov/wip-krbd-58-options-octopus

octopus: rbd: recognize crush_location, read_from_replica and compression_hint map options

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

doc: Update nfs-ganesha package requirements

Fixes: https://tracker.ceph.com/issues/46449
Signed-off-by: Varsha Rao <varao@redhat.com>
(cherry picked from commit 16900045f4729ad6044ba60bd84dc5d1c0591c14)

qa/suites/krbd: turn on balanced reads for the fsx subsuite

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
(cherry picked from commit d15e0cad1a6c1912820c43463cd8f8576968a870)

test/librbd/fsx: respect rbd_default_map_options in krbd_open()

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
(cherry picked from commit b29dfc4624f3b8286d345ec62894fa70c5f6cf9b)

rbd: recognize compression_hint map option

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
(cherry picked from commit 4983eabd6fe75bb6f669d9d18388349f342a5d27)

rbd: recognize crush_location and read_from_replica map options

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
(cherry picked from commit e3874a25f0e828e2a6bd8735dbeb14197008896c)

Merge pull request #36050 from smithfarm/wip-46307-octopus

octopus: tests: unittest_lockdep: skip lockdep test if CEPH_DEBUG_MUTEX is not defined

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

test/unittest_lockdep: skip lockdep test if CEPH_DEBUG_MUTEX is not defined

this change addresses a regression introduced by
2dc50b5f91476cf10429dbcdd52b00e32f599d5c

Fixes: https://tracker.ceph.com/issues/46267
Signed-off-by: Kefu Chai <kchai@redhat.com>
(cherry picked from commit fb74d5443a289c5235210f4c0f1500f2a90cb047)