16.2.12

Signed-off-by: Ceph Release Team <ceph-maintainers@ceph.io>

16.2.12

Signed-off-by: Ceph Release Team <ceph-maintainers@ceph.io>

Merge branch 'cv-bkp-50473-pacific' of https://github.com/guits/ceph into pacific-release

ceph-volume/tests: fix an issue with rpm

Typical error seen in the CI:

```
error: /var/cache/dnf/baseos-00fe51d07def85f0/packages/kernel-core-4.18.0-483.el8.x86_64.rpm: signature hdr data: BAD, no. of bytes(459772) out of range
```

Upgrading `rpm` fixes this issue.

Signed-off-by: Guillaume Abrioux <gabrioux@ibm.com>
(cherry picked from commit 05aa334b96ed8a8c8497047bc26da64ee3a3d1fa)

ceph-volume: add bluefs_buffered_io: false in functional tests

This is a workaround to make dmcrypt scenarios pass the CI.

Signed-off-by: Guillaume Abrioux <gabrioux@ibm.com>
(cherry picked from commit 491e88ede56b51bc425ba5e38adf6eb74ad5196e)

ceph-volume: update the OS before deploying Ceph

ceph-volume tests are failing, OSDs never get up and running.
For some reason, updating the OS early in the testing workflow
addresses that issue in the CI.

-- to be continued ... --

Signed-off-by: Guillaume Abrioux <gabrioux@ibm.com>
(cherry picked from commit 663c914c59873f433a5512aca0ebc4b4a7ef1fd3)

Merge pull request #50914 from rhcs-dashboard/wip-59347-pacific

pacific: mgr/dashboard: ignore the rules 400 error in dashboard kcli e2e

Reviewed-by: Pegonzal <NOT@FOUND>

Merge pull request #50121 from rhcs-dashboard/wip-57830-pacific

pacific: mgr/dashboard: fix weird data in osd details

Reviewed-by: Pegonzal <NOT@FOUND>

Merge pull request #50968 from zdover23/wip-doc-2023-04-10-backport-50828-to-pacific

pacific: doc/rados/ops: edit health checks.rst (5 of x)

Reviewed-by: Cole Mitchell <cole.mitchell.ceph@gmail.com>

Merge pull request #50878 from guits/wip-59311-pacific

pacific: ceph-volume: fix issue with fast device allocs when there are multiple PVs per VG

Merge pull request #50971 from zdover23/wip-doc-2023-04-10-backport-50829-to-pacific

pacific: doc/rados/ops: edit health-checks.rst (6 of x)

doc/rados/ops: edit health-checks.rst (6 of x)

Edit docs/rados/operations/health-checks.rst (6 of x).

Follows https://github.com/ceph/ceph/pull/50828.

https://tracker.ceph.com/issues/58485

Co-authored-by: Anthony D'Atri <anthony.datri@gmail.com>
Signed-off-by: Zac Dover <zac.dover@proton.me>
(cherry picked from commit c59bc152ea746a8bfc5795dbeb68b2ef6e511b1f)

doc/rados/ops: edit health checks.rst (5 of x)

Edit docs/rados/operations/health-checks.rst (5 of x).

Follows https://github.com/ceph/ceph/pull/50827

https://tracker.ceph.com/issues/58485

Co-authored-by: Anthony D'Atri <anthony.datri@gmail.com>
Signed-off-by: Zac Dover <zac.dover@proton.me>
(cherry picked from commit cb8ec5b5ab8f5fb25c8f2c6404447dabe4ebd1ad)

Merge pull request #50961 from zdover23/wip-doc-2023-04-09-backport-50958-to-pacific

pacific: doc/rados/ops: add hyphen to mon-osd-pg.rst

doc/rados/ops: add hyphen to mon-osd-pg.rst

Remove confusing parentheses and add a clearer (as compared to the
parentheses) hyphen (actually an em-dash, or at least it is intended
to be an em-dash) to doc/rados/operations/monitoring-osd-pg.rst

Co-authored-by: Anthony D'Atri <anthony.datri@gmail.com>
Signed-off-by: Zac Dover <zac.dover@proton.me>
(cherry picked from commit 0c965c18d0e6ab1461b5fad42d481f25e4207940)

Merge pull request #50957 from zdover23/wip-doc-2023-04-09-backport-50827-to-pacific

pacific: doc/rados/ops: edit health-checks.rst (4 of x)

Reviewed-by: Anthony D'Atri <anthony.datri@gmail.com>

Merge pull request #50954 from zdover23/wip-doc-2023-04-09-backport-50826-to-pacific

pacific: doc/rados/ops: edit health-checks.rst (3 of x)

Reviewed-by: Anthony D'Atri <anthony.datri@gmail.com>

Merge pull request #50951 from zdover23/wip-doc-2023-04-09-backport-50907-to-pacific

pacific: doc/rados/config: edit auth-config-ref

Reviewed-by: Anthony D'Atri <anthony.datri@gmail.com>

doc/rados/ops: edit health-checks.rst (4 of x)

Edit docs/rados/operations/health-checks.rst (4 of x).

Follows https://github.com/ceph/ceph/pull/50826.

https://tracker.ceph.com/issues/58485

Co-authored-by: Anthony D'Atri <anthony.datri@gmail.com>
Signed-off-by: Zac Dover <zac.dover@proton.me>
(cherry picked from commit 31df75356f2e49e0da6e0f4a6ec498e3609bd39e)

doc/rados/ops: edit health-checks.rst (3 of x)

Edit docs/rados/operations/health-checks.rst (3 of x).

Follows https://github.com/ceph/ceph/pull/50825

https://tracker.ceph.com/issues/58485

Co-authored-by: Anthony D'Atri <anthony.datri@gmail.com>
Signed-off-by: Zac Dover <zac.dover@proton.me>
(cherry picked from commit e02ddb24e583b5bedd4056ba0871a9427b581bfe)

doc/rados/config: edit auth-config-ref

Line edit doc/rados/config/auth-config-ref.rst.

https://tracker.ceph.com/issues/58485

Co-authored-by: Anthony D'Atri <anthony.datri@gmail.com>
Signed-off-by: Zac Dover <zac.dover@proton.me>
(cherry picked from commit 3d0acf9372a35ef48be47efe68192ec389c17ccb)

Merge pull request #50947 from zdover23/wip-doc-2023-04-08-backport-50863-to-pacific

pacific: doc/rados/ops: edit monitoring-osd-pg.rst (2 of x)

Reviewed-by: Cole Mitchell <cole.mitchell.ceph@gmail.com>

Merge pull request #50944 from zdover23/wip-doc-2023-04-08-backport-50795-to-pacific

pacific: doc/rados: line-edit common.rst

Reviewed-by: Anthony D'Atri <anthony.datri@gmail.com>

doc/rados/ops: edit monitoring-osd-pg.rst (2 of x)

Line-edit monitoring-osd-pg.rst (2 of x).

https://tracker.ceph.com/issues/58485

Co-authored-by: Anthony D'Atri <anthony.datri@gmail.com>
Signed-off-by: Zac Dover <zac.dover@proton.me>
(cherry picked from commit 270e2fd730dad19d970d73ea19c6872defa9fce4)

doc/rados: line-edit common.rst

Edit syntax and semantics in doc/configuration/common.rst.

Signed-off-by: Zac Dover <zac.dover@proton.me>
(cherry picked from commit 6c9f226742e951de6c40627271ba6d6cabba4535)

Merge pull request #50850 from zdover23/wip-doc-2023-04-04-backport-50834-to-pacific

pacific: doc/rados: edit ops/monitoring.rst (2 of 3)

Reviewed-by: Cole Mitchell <cole.mitchell.ceph@gmail.com>

Merge pull request #50832 from zdover23/wip-doc-2023-04-02-backport-50818-to-pacific

pacific: doc/cephfs: Improve fs-volumes.rst

Reviewed-by: Cole Mitchell <cole.mitchell.ceph@gmail.com>

Merge pull request #50935 from zdover23/wip-doc-2023-04-07-backport-50928-to-pacific

pacific: doc/rados/ops: line-edit operating.rst

doc/rados/ops: line-edit operating.rst

Line-edit doc/rados/operations/operating.rst.

https://tracker.ceph.com/issues/58485

Signed-off-by: Zac Dover <zac.dover@proton.me>
(cherry picked from commit bb8fc0068369369fd213062590738e38c9f4dc63)

doc/cephfs: Improve fs-volumes.rst

Signed-off-by: Anthony D'Atri <anthonyeleven@users.noreply.github.com>
(cherry picked from commit c8e1f3ac3db513004fbfbb0049757ddc2bca291a)

doc/rados: edit ops/monitoring.rst (2 of 3)

Line-edit the second third of doc/rados/operations/monitoring.rst.

Follows https://github.com/ceph/ceph/pull/50670.

https://tracker.ceph.com/issues/58485

Co-authored-by: Anthony D'Atri <anthony.datri@gmail.com>
Signed-off-by: Zac Dover <zac.dover@proton.me>
(cherry picked from commit 41684ebd33b5c9fe707f5a33b27c55ed29cd5ede)

Merge PR #50923 into pacific

* refs/pull/50923/head:
qa: load file system info if not created

Reviewed-by: Venky Shankar <vshankar@redhat.com>

Merge PR #50860 into pacific

* refs/pull/50860/head:
qa: simplify and use correct recovery procedure
doc: update alternate meta pool recovery
tools/cephfs/DataScan: add debugging for directory injection

Reviewed-by: Venky Shankar <vshankar@redhat.com>

Merge PR #50784 into pacific

* refs/pull/50784/head:
tools/cephfs: include lost+found in scan_links

Reviewed-by: Venky Shankar <vshankar@redhat.com>

Merge PR #50781 into pacific

* refs/pull/50781/head:
mds: add config to decide whether to mark dentry bad
qa: add missing scan_links step for data scan recovery
qa/tasks/cephfs: test damage to dentry's first is caught
qa/tasks/cephfs: use rank_asok and allow specifying rank
qa/tasks: allow specifying timeout command prefix to ceph
mds: provide test configs for creating first corruption
mds: catch damage to dentry's first field
mds: add debugging for pre_cow_old_inode
mds: cleanup code
mds: check for some dentry damage in scrub
mds: remove unused method
mds: note damaged dentry with first gt last
mds: cluster log scrub failure for dirfrag
mds: mark dirfrag good if repaired
mds: only dump past_parent_snap if non-empty

Reviewed-by: Venky Shankar <vshankar@redhat.com>

Merge PR #50775 into pacific

* refs/pull/50775/head:
qa: ignore expected scrub error

Reviewed-by: Venky Shankar <vshankar@redhat.com>

Merge PR #50760 into pacific

* refs/pull/50760/head:
qa: lengthen health warning wait

Reviewed-by: Venky Shankar <vshankar@redhat.com>

Merge PR #50757 into pacific

* refs/pull/50757/head:
qa: ignore MDS_TRIM warnings when osd thrashing

Reviewed-by: Venky Shankar <vshankar@redhat.com>

Merge pull request #50913 from zdover23/wip-doc-2023-04-06-backport-50825-to-pacific

pacific: doc/rados/ops: edit health-checks.rst (2 of x)

qa: load file system info if not created

Mounts may create a Filesystem object which does not create the file
system. This causes self.id to be Null when we try to wait for daemons
to be up. Load it just in time!

Fixes: https://tracker.ceph.com/issues/59332
Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
(cherry picked from commit 69565dbbf124e29f30dae989eae41eada161a4b4)

mgr/dashboard: ignore the rules 400 error in dashboard kcli e2e

Fixes: https://tracker.ceph.com/issues/58722
Signed-off-by: Nizamudeen A <nia@redhat.com>
(cherry picked from commit a9634e902afcd2deda490df08080686d4dc110e0)

Merge pull request #50123 from rhcs-dashboard/wip-57433-pacific

pacific: mgr/dashboard: fix "can't read .ssh/known_hosts: No such file or directory

Reviewed-by: Pegonzal <NOT@FOUND>

doc/rados/ops: edit health-checks.rst (2 of x)

Edit docs/rados/operations/health-checks.rst (2 of x). PR#50674, the PR
that immediately precedes this PR in the series of PRs that line-edit
health-checks.rst, wrongly identified this series as having five
sections. This has been rectified by using the "2 of x" formulation.

Follows https://github.com/ceph/ceph/pull/50674

https://tracker.ceph.com/issues/58485

Signed-off-by: Zac Dover <zac.dover@proton.me>
(cherry picked from commit 0565f59698e9d00adf5a4667ac6e84e97d31b554)

Merge branch 'wip-59311-pacific' of https://github.com/guits/ceph into pacific-release

ceph-volume: add test case to reproduce bug in get_physical_fast_allocs

Adds a test case to reproduce a bug with get_physical_fast_allocs for
clusters that have multiple fast device PVs in a single VG (deployed
prior to v15.2.8). Also fixes other test cases for this function
to more accurately represent reality.

Signed-off-by: Cory Snyder <csnyder@1111systems.com>
(cherry picked from commit 02592cb0a5970f861d25a521579c30639d10b007)

ceph-volume: fix issue with fast device allocs when there are multiple PVs per VG

Fixes a regression with fast device allocations when there are multiple PVs
per VG. This is the case for clusters that were deployed prior to v15.2.8.

Fixes: https://tracker.ceph.com/issues/58857
Signed-off-by: Cory Snyder <csnyder@1111systems.com>
(cherry picked from commit efcf71be18eb25be10a574d54b70229753538664)

ceph-volume: legacy_encrypted() shouldn't call lsblk() when device is 'tmpfs'

When the device variable is rebound and become 'tmpfs', lsblk() shouldn't be called.

Fixes: https://tracker.ceph.com/issues/58784
Signed-off-by: Guillaume Abrioux <gabrioux@ibm.com>
(cherry picked from commit c0369fbecee1a4d71004ee5e72ee0e2526d0fcf2)
(cherry picked from commit 797cc5a9056fe10e222784fdfb068e8dfd35b865)

ceph-volume: do not raise RuntimeError in util.lsblk

Otherwise it makes `simple scan` calls fail.

```
  File "/usr/lib/python3.6/site-packages/ceph_volume/util/disk.py", line 246, in lsblk
    raise RuntimeError(f"{device} not found is lsblk report")
RuntimeError: /var/lib/ceph/osd/ceph-0/ not found is lsblk report
```

Fixes: https://tracker.ceph.com/issues/58655
Signed-off-by: Guillaume Abrioux <gabrioux@ibm.com>
(cherry picked from commit 4f19a0f4e1726466968d610da3b0f753981e44a7)
(cherry picked from commit 259a3d554f53c549b135570f2054c3b5c517f993)

ceph-volume: fix a bug in lsblk_all()

Rook has a specific use case where devices are copied in /mnt

If the basename (in /mnt) is different from the original device name,
then the current logic can't match it.

The idea is to append the device to the `lsblk` command and return
the result.

Fixes: https://tracker.ceph.com/issues/58137
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
(cherry picked from commit dce4d469f66124babe97dd3651f67d55e805088c)
(cherry picked from commit 658472614032d8a8f5e8ea5ddb8759e762c69bb2)

ceph-volume: fix regression in activate

bea9f4b643c introduced a regression that makes the activate process
take a very long time to complete.
`_get_bluestore_info()` which calls `ceph-bluestore-tool` binary via
subprocess is called in an exponential way while this is not needed.

Fixes: https://tracker.ceph.com/issues/57627
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
(cherry picked from commit 8d7423c3e75afbe111c91e699ef3cb1c0beee61b)
(cherry picked from commit d322bb891dbaf420dd1d1d32d4ce3e8582bd715b)

ceph-volume: fix a bug in get_all_devices_vgs()

When a physical volume isn't a member of any volume group,
the function `get_all_devices_vgs()` in `ceph_volume.api.lvm`
shouldn't try to return an `VolumeGroup()` object against this
device.

Fixes: https://tracker.ceph.com/issues/57918
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
(cherry picked from commit 770cd09412f933eeb93426a8acfc4f4666c99024)
(cherry picked from commit bd9b1c8eb5d0f1708d7fae50e85d6bf0b52f38c7)

Merge pull request #50036 from batrick/i52247

pacific: libcephsqlite: CheckReservedLock the result will always be zero

Reviewed-by: Laura Flores <lflores@redhat.com>

Merge pull request #50092 from petrutlucian94/wip-58703-pacific

pacific: os/memstore: Fix memory leak

Reviewed-by: Ilya Dryomov <idryomov@redhat.com>

Merge pull request #50072 from ifed01/wip-ifed-more-pin-fix-pac

pacific: os/bluestore: fix onode ref counting

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

Merge pull request #49937 from NitzanMordhai/wip-58613-pacific

pacific: Wip nitzan pglog ec getattr error

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

Merge pull request #49830 from tserong/wip-58533-pacific

pacific: ceph.spec.in: Replace %usrmerged macro with regular version check

Reviewed-by: Ilya Dryomov <idryomov@redhat.com>

mgr/dashboard: auto-coloring-badges-component

Signed-off-by: Pedro Gonzalez Gomez <pegonzal@redhat.com>
(cherry picked from commit 29f510a522b2694903d095ac0ec45ee23c474121)

ceph-volume: fix issue with fast device allocs when there are multiple PVs per VG

Fixes a regression with fast device allocations when there are multiple PVs
per VG. This is the case for clusters that were deployed prior to v15.2.8.

Fixes: https://tracker.ceph.com/issues/58857
Signed-off-by: Cory Snyder <csnyder@1111systems.com>
(cherry picked from commit efcf71be18eb25be10a574d54b70229753538664)

ceph-volume: add test case to reproduce bug in get_physical_fast_allocs

Adds a test case to reproduce a bug with get_physical_fast_allocs for
clusters that have multiple fast device PVs in a single VG (deployed
prior to v15.2.8). Also fixes other test cases for this function
to more accurately represent reality.

Signed-off-by: Cory Snyder <csnyder@1111systems.com>
(cherry picked from commit 02592cb0a5970f861d25a521579c30639d10b007)

Merge pull request #50870 from rhcs-dashboard/wip-59321-pacific

pacific: mgr/dashboard: fix displaying mirror image progress

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

Merge pull request #47725 from rhcs-dashboard/wip-57217-pacific

pacific: mgr/dashboard: Hide maintenance option on expand cluster

Reviewed-by: Aashish Sharma <aasharma@redhat.com>
Reviewed-by: Avan Thakkar <athakkar@redhat.com>
Reviewed-by: Pere Diaz Bou <pdiazbou@redhat.com>

Merge pull request #50866 from zdover23/wip-doc-2023-04-05-backport-50855-to-pacific

pacific: doc/rados/ops: edit monitoring-osd-pg.rst (1 of x)

mgr/dashboard: fix displaying mirror image progress

Signed-off-by: Pere Diaz Bou <pere-altea@hotmail.com>
(cherry picked from commit 80960387fbb6d49c3681f4c9e249c441a891462a)

doc/rados/ops: edit monitoring-osd-pg.rst (1 of x)

Line-edit doc/rados/operations/monitoring-osd-pg.rst (1 of x).

https://tracker.ceph.com/issues/58485

Co-authored-by: Anthony D'Atri <anthony.datri@gmail.com>
Signed-off-by: Zac Dover <zac.dover@proton.me>
(cherry picked from commit bd213c244e8a387c98358ac41f4802e102e89c63)

Merge pull request #50289 from idryomov/wip-58833-pacific

pacific: librbd: call apply_changes() after setting librados_thread_count

Reviewed-by: Christopher Hoffman <choffman@redhat.com>

qa: simplify and use correct recovery procedure

Notably, this recovery procedure was missing scan_links.

Also, the test was oddly trying to recover the real file system in
addition to the recovery file system. I've removed that unnecessary
recovery.

Fixes: https://tracker.ceph.com/issues/57598
Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
(cherry picked from commit 8512bbc01525b77583a87dfc2149fb1dec072978)

Conflicts:
qa/tasks/cephfs/filesystem.py: trivial

doc: update alternate meta pool recovery

Streamline, clarify, and use modern commands (notably: --recover).

Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
(cherry picked from commit da19a128ee4a68299d4899de91701944720f53ae)

tools/cephfs/DataScan: add debugging for directory injection

Continuation of 16fc2c392b0a4de13dbe4dfcc7dad01d8e58ce40.

Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
(cherry picked from commit 94d2e24b5cb2bbbfb2d55f6cc538d448091e3d4e)

Merge pull request #50854 from zdover23/wip-doc-2023-04-04-backport-50851-to-pacific

pacific: doc/rados/ops: add ceph-medic documentation

doc/rados/ops: add ceph-medic documentation

Add a link to the ceph-medic documentation to
doc/rados/operations/monitoring.rst.

Signed-off-by: Zac Dover <zac.dover@proton.me>
(cherry picked from commit b2a409da26b508e5a51e3becbce16b7ecaffcc1d)

Merge pull request #50824 from zdover23/wip-doc-2023-04-01-backport-50670-to-pacific

pacific: doc/rados: edit ops/monitoring.rst (1 of 3)

Reviewed-by: Cole Mitchell <cole.mitchell.ceph@gmail.com>

doc/rados: edit ops/monitoring.rst (1 of 3)

Line-edit the first third of doc/rados/operations/monitoring.rst.

https://tracker.ceph.com/issues/58485

Signed-off-by: Zac Dover <zac.dover@proton.me>
(cherry picked from commit bc5f404a7876f887d0c7a2dedee5391f66ab0505)

Merge pull request #50801 from zdover23/wip-doc-2023-03-31-backport-45354-to-pacific

pacific: doc/msgr2: update dual stack status

Merge pull request #50804 from zdover23/wip-doc-2023-03-31-backport-50554-to-pacific

pacific: doc/rados: line-edit pg-repair.rst

Merge pull request #50798 from zdover23/wip-doc-2023-03-31-backport-50674-to-pacific

pacific: doc/rados/ops: edit health-checks.rst (1 of x)

Reviewed-by: Anthony D'Atri <anthony.datri@gmail.com>

doc/rados: line-edit pg-repair.rst

Line edit pg-repair.rst.

Signed-off-by: Zac Dover <zac.dover@proton.me>
(cherry picked from commit b7c19dac2dfdcbc50515e83d952fb25070496c1c)

doc/msgr2: update dual stack status

Signed-off-by: Dan van der Ster <daniel.vanderster@cern.ch>
(cherry picked from commit c7eaeb4dff1fca967046bcd9b53548eabd18012e)

doc/rados/ops: edit health-checks.rst (1 of x)

Edit docs/rados/operations/health-checks.rst (1 of x).

https://tracker.ceph.com/issues/58485

Signed-off-by: Zac Dover <zac.dover@proton.me>
(cherry picked from commit cb1e20f879edb82ae619693a4f1f5e88fd0593dd)
(cherry picked from commit bfeadab4cacfea984b24be32f74aae29d750c148)

Merge pull request #50771 from zdover23/wip-doc-2023-03-30-backport-50742-to-pacific

pacific: doc/start: update "notify us" section

doc/start: update "notify us" section

Rewrite the "Notify Us" section in doc/start/documenting-ceph.rst so
that it is written in English sentences.

(cherry picked from commit 613ece24777c4b222054fe9a6d82ba51949c45ea)
Signed-off-by: Zac Dover <zac.dover@proton.me>

qa: lengthen health warning wait

It's just a little too short!

Fixes: https://tracker.ceph.com/issues/52995
Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
(cherry picked from commit c8810e46e8f28ed204ef2d39f552f96f4a6cfe39)

Conflicts:
qa/tasks/cephfs/test_failover.py

mds: add config to decide whether to mark dentry bad

So admin can restore access to files if necessary.

Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
(cherry picked from commit 7ffa065a03316b67b6b6d48692b4b181f93ddbda)

qa: add missing scan_links step for data scan recovery

Without, the first field remains corrupt (HEAD).

Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
(cherry picked from commit 1b4d9077af826d1931b91938fd49e7113e8d3840)

Conflicts:
qa/tasks/cephfs/test_data_scan.py

qa/tasks/cephfs: test damage to dentry's first is caught

Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
(cherry picked from commit 293b90fe63665666ca2de8b65c38f987a4d0b2a6)

qa/tasks/cephfs: use rank_asok and allow specifying rank

This will use the more efficient:

    ceph tell mds.<fsname>:<rank> ...

Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
(cherry picked from commit 3ba17390e92c21a3eba7337a2da0df79f6bf49ca)

qa/tasks: allow specifying timeout command prefix to ceph

Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
(cherry picked from commit dbf1a402c9626bcb2d1b3363f871c569860a0921)

Conflicts:
qa/tasks/ceph_manager.py

Merge pull request #50789 from zdover23/wip-doc-2023-03-30-backport-50764-to-pacific

pacific: doc/start: format procedure in documenting-ceph

Reviewed-by: Anthony D'Atri <anthony.datri@gmail.com>

doc/start: format procedure in documenting-ceph

Organize three steps in a procedure into one step with three sub-steps.

Signed-off-by: Zac Dover <zac.dover@proton.me>
(cherry picked from commit 6f3cd78f6116f3e19a8f6c5e0ebda816067a334e)

tools/cephfs: include lost+found in scan_links

Otherwise, any injected dentries have incorrect first snapids.

Fixes: https://tracker.ceph.com/issues/59183
Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
(cherry picked from commit d8b6d451847cda2b3b06ed630a14306cd20bf9b8)

mds: provide test configs for creating first corruption

Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
(cherry picked from commit fe258dc54bbeee5913289cad08634b5ff1d737ba)

Conflicts:
src/mds/MDSRank.cc
src/mds/MDSRank.h
src/mds/Server.cc

mds: catch damage to dentry's first field

When possible. Abort the MDS before it can be written to the
journal/directory.

This is part of a series to address corruption first observed in [1].
How the corruption is introduced is yet unknown.

[1] https://tracker.ceph.com/issues/38452#note-10

Fixes: http://tracker.ceph.com/issues/58482
Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
(cherry picked from commit 03407528f95fe60e5af4062c3caa3688b8b31530)

Conflicts:
src/common/options/mds.yaml.in
src/mds/CDir.cc

mds: add debugging for pre_cow_old_inode

Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
(cherry picked from commit c9d36056c7d515ad879d589938f888f9f2787973)

mds: cleanup code

Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
(cherry picked from commit 17d826551cd9466b4c5457e668d581baa9252c91)

mds: check for some dentry damage in scrub

Fixes: https://tracker.ceph.com/issues/57091
Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
(cherry picked from commit 14712f96294ad0915951af151e4919ebf6c40cbb)

mds: remove unused method

Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
(cherry picked from commit 104ae798c68c95faf904adb498aae35063fe04a0)

mds: note damaged dentry with first gt last

Only when loaded in cache.

Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
(cherry picked from commit 956eff7e02498511b597aadb85c686b78c44ea8a)

mds: cluster log scrub failure for dirfrag

Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
(cherry picked from commit 40f483ed457a24cfdab343875ab98ca1240850a6)

mds: mark dirfrag good if repaired

Otherwise, we get wrong scrub errors after repair.

Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
(cherry picked from commit 7d6199a226f7e54b9580cbf05d97114068377b4d)

mds: only dump past_parent_snap if non-empty

Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
(cherry picked from commit b1a169b4f319c92ed0e57f68e224572fb509619c)

qa: ignore expected scrub error

The test_orphan_scan test deliberately removes a dentry which will cause
rstat damage. Ignore it.

Fixes: https://tracker.ceph.com/issues/57657
Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
(cherry picked from commit e930a8b22108917f195ce5e106b1c70cabef9c91)