mgr/dashboard: disable Promql test in ARM

Temporarily disable this test while debugging the issue (since https://github.com/ceph/ceph/pull/43669
originally passed the ARM check).

Fixes: https://tracker.ceph.com/issues/53451
Signed-off-by: Ernesto Puerta <epuertat@redhat.com>

Merge pull request #44009 from cbodley/wip-qa-cls-rgw-gc

qa/rgw: run ceph_test_cls_rgw_gc in rgw/verify suite

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

Merge pull request #42905 from sebastian-philipp/service_spec_no_redundant_placement

python-common: improve OSD spec error messages

Reviewed-by: Adam King <adking@redhat.com>
Reviewed-by: Alfonso Martínez <almartin@redhat.com>
Reviewed-by: Michael Fritch <mfritch@suse.com>

Merge pull request #44218 from guits/guits-issue-44356

ceph-volume: fix error 'KeyError' with inventory

Merge pull request #44219 from guits/guits-issue-53425

ceph-volume: fix tags dict output in `lvm list`

Merge pull request #44281 from athanatos/sjust/wip-53555

crimson/os/seastore: index lba pins atomically with addition to cache

Reviewed-by: Xuehan Xu <xuxuehan@360.cn>
Reviewed-by: Yingxin Cheng <yingxin.cheng@intel.com>

Merge pull request #44225 from liewegas/fix-53506

osd/OSDMapMapping: fix spurious threadpool timeout errors

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

Merge pull request #44025 from ljflores/wip-remove-aggregated-perf-data

mgr/telemetry: remove aggregated perf metrics from the perf channel

Reviewed-by: Nizamudeen A <nia@redhat.com>
Reviewed-by: Yaarit Hatuka <yaarit@redhat.com>
Reviewed-by: Alfonso Martínez <almartin@redhat.com>

Merge pull request #43612 from adamemerson/wip-unleveling

build: Remove LevelDB support

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

crimson/os/seastore/cache: init extents prior to read

Thus should ensure that any captured members of extent_init_func are
still valid at the cost of not being able to access the contents of the
extent at invocation time.  With this, we should be able to rely on any
logical extents/lba extents in the cache having validly initialized lba
pins.

Fixes: https://tracker.ceph.com/issues/53555
Signed-off-by: Samuel Just <sjust@redhat.com>

crimson/os/seatore/.../lba_btree: update get_*_node to add_pin without reading node contents

This will allow us to do add_pin before we perform the actual extent read.

Signed-off-by: Samuel Just <sjust@redhat.com>

crimson/os/seastore: pass depth/begin/end to get_*_node

We'll need this to populate the pin fields prior to read.

Signed-off-by: Samuel Just <sjust@redhat.com>

crimson/os/seastore/transaction_manager: clarify that init lambda only runs on new extents

Signed-off-by: Samuel Just <sjust@redhat.com>

crimson/os/seastore/cache: remove unnecessary forward

Signed-off-by: Samuel Just <sjust@redhat.com>

Merge pull request #44262 from cbodley/wip-53325

test/rgw: disable cls_rgw_gc test cases with defer_gc()

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

Merge pull request #44206 from linuxbox2/wip-rgwlc-9969

rgwlc:  remove magic debug blocks for clearing stale lc entries

Merge pull request #43991 from rhcs-dashboard/introduce-haroxy-metrics-rgw

mgr/dashboard: introduce HAProxy metrics for RGW

Reviewed-by: Alfonso Martínez <almartin@redhat.com>
Reviewed-by: Ernesto Puerta <epuertat@redhat.com>
Reviewed-by: Matt Benjamin <mbenjami@redhat.com>
Reviewed-by: Pere Diaz Bou <pdiazbou@redhat.com>

Merge pull request #43258 from soumyakoduri/wip-skoduri-dbstore-mp

rgw/dbstore: Multipart upload APIs

Merge pull request #42104 from linuxbox2/wip-rgwadminops-fsid

rgw: expose RADOS cluster_fsid via adminops

rgw:adminops: remove "import json" from radosgw_admin_rest.py

This is perhaps erring a bit on the side of cosmetic fixes.

Signed-off-by: Matt Benjamin <mbenjamin@redhat.com>

Merge pull request #44220 from vshankar/wip-53487

qa: exclude `nofallback` mount option when using v1-style syntax

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

qa/rgw: run ceph_test_cls_rgw_gc in rgw/verify suite

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

test/rgw: disable cls_rgw_gc test cases with defer_gc()

Fixes: https://tracker.ceph.com/issues/53325
Signed-off-by: Casey Bodley <cbodley@redhat.com>

rgw:adminops: slightly generalize /info

Adds a get_name() method to rgw::sal::Store, by which each store
returns its unique name in lowercase.

Signed-off-by: Matt Benjamin <mbenjamin@redhat.com>

rgw:sal: implement (placeholder) get_cluster_fsid in DBStore

Signed-off-by: Matt Benjamin <mbenjamin@redhat.com>

rgw:adminops: add test case for 'info' section

Add 'info' section test case to the radosgw_admin_test.py qa
task.

Signed-off-by: Matt Benjamin <mbenjamin@redhat.com>

rgw: introduce new /{admin}/info api resource

The new resource returns an array of informational data, which
currently consists of the RADOS cluster fsid as "cluster_fsid."

Fixes: https://tracker.ceph.com/issues/51432
Signed-off-by: Matt Benjamin <mbenjamin@redhat.com>

rgw:sal: expose cluster fsid in RGWStore

Adds a get_cluster_id() method to RGWStore, provided librados
cluster_fsid in the RADOS implementation.

Signed-off-by: Matt Benjamin <mbenjamin@redhat.com>

Merge pull request #44001 from votdev/issue_53317_fix_proxy_conf

mgr/dashboard: API docs UI does not work with Angular dev server

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

Merge pull request #44237 from pritha-srivastava/wip-rgw-abort-multipart-bypass-gc

rgw: deleting objects inline in case bypass_gc is specified

mgr/dashboard: introduce HAProxy metrics for RGW

Fixes: https://tracker.ceph.com/issues/53311
Signed-off-by: Avan Thakkar <athakkar@redhat.com>

qa: exclude `nofallback` mount option when using v1-style syntax

Otherwise, certain upgrade tests fail which install pacific
or earlier releases since the mount helper does not understand
this mount option, thereby passing it to the kernel which would
does not handle this config causing mount to fail in tests.

Note that this mount config is only used during teuthology tests
to catch v2-style syntax implementation bugs in the kernel.

Fixes: http://tracker.ceph.com/issues/53487
Signed-off-by: Venky Shankar <vshankar@redhat.com>

Merge pull request #44127 from cyx1231st/wip-seastore-merge-journal-headers-2

crimson/os/seastore: make JournalSubmitter aware of header merging

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

crimson/os/seastore/journal: make JournalSubmitter aware of header merging

Introduce an option seastore_journal_batch_preferred_fullness for
JournalSubmitter to control the preferred fullness to trigger batch
flushes.

Signed-off-by: Yingxin Cheng <yingxin.cheng@intel.com>

crimson/os/seastore: refactor get_encoded_length_after()

To return a more informative record_group_size_t.

Signed-off-by: Yingxin Cheng <yingxin.cheng@intel.com>

crimson/os/seastore: debug print journal write details

Signed-off-by: Yingxin Cheng <yingxin.cheng@intel.com>

crimson/os/seastore: measure records write in journal

Signed-off-by: Yingxin Cheng <yingxin.cheng@intel.com>

crimson/os/seastore: fix record metrics in cache

Signed-off-by: Yingxin Cheng <yingxin.cheng@intel.com>

crimson/os/seastore/cache: misc cleanup

Signed-off-by: Yingxin Cheng <yingxin.cheng@intel.com>

crimson/os/seastore: add debug logs to print transaction name

Signed-off-by: Yingxin Cheng <yingxin.cheng@intel.com>

crimson/os/seastore: don't print detail if the extent is not ready

Signed-off-by: Yingxin Cheng <yingxin.cheng@intel.com>

crimson/os/seastore: print extent length

Signed-off-by: Yingxin Cheng <yingxin.cheng@intel.com>

crimson/os/seastore: move implementation to seastore_types.cc

Signed-off-by: Yingxin Cheng <yingxin.cheng@intel.com>

crimson/os/seastore: merge records metadata if they are grouped

Signed-off-by: Yingxin Cheng <yingxin.cheng@intel.com>

crimson/os/seastore: mark out empty transactions

TODO: avoid write if the transaction is empty.

Signed-off-by: Yingxin Cheng <yingxin.cheng@intel.com>

crimson/os/seastore: refactor, introduce record_t and record_group_t with sizes

Signed-off-by: Yingxin Cheng <yingxin.cheng@intel.com>

crimson/os/seastore: scan records based on record_locator_t

Record may not have its own base if headers are merged.

Signed-off-by: Yingxin Cheng <yingxin.cheng@intel.com>

crimson/os/seastore: add more checks when read record_header_t

Signed-off-by: Yingxin Cheng <yingxin.cheng@intel.com>

crimson/os/seastore: misc cleanup and reformat

Signed-off-by: Yingxin Cheng <yingxin.cheng@intel.com>

crimson/os/seastore: add logs in ExtentReader

Signed-off-by: Yingxin Cheng <yingxin.cheng@intel.com>

Merge pull request #43919 from ronen-fr/wip-rf-test-nodeep

osd/scrub (& qa/standalone): test for scrub behavior when no-scrub is set but no-deep-scrub is not

Reviewed-by: Neha Ojha <nojha@redhat.com>
Reviewed-by: Matan Breizman <Matan.Brz@gmail.com>

Merge pull request #43305 from heylinn/ceph_rundir_sysvinit

init-ceph: create /var/run/ceph for sysvinit

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

Merge pull request #44090 from sseshasa/wip-fix-require-osd-release

osd/OSDMap: Add health warning if 'require-osd-release' != current release

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

Merge pull request #44244 from cyx1231st/wip-seastore-refine-metrics

crimson/os/seastore: refine transaction metrics

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

Merge pull request #44242 from liu-chunmei/crimson-fix-heartbeat-addrs

crimson/osd: fix heartbeat front and back blank ip

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

Merge pull request #44098 from ifed01/wip-ifed-dump-alloc-unit

os/bluestore: dump bluestore/bluefs alloc unit sizes with perf dump

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

Merge pull request #43840 from ifed01/wip-ifed-verbose-open-col

osd,bluestore: gracefully handle a failure during meta collection load

Reviewed-by: jdurgin@redhat.com
Reviewed-by: nojha@redhat.com

Merge pull request #39691 from aclamk/wip-bs-compression-blob-64k

os/bluestore: Set new compression blob size to 64K

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

Merge pull request #44213 from zdover23/wip-doc-2021-12-05-hardware-recommendations-removing-journal

doc/start: remove journal info from hardware recs

Reviewed-by: Dan van der Ster <daniel.vanderster@cern.ch>

rgw/dbstore: Multipart upload APIs

For multipart upload processing, below is the method applied -

MultipartUpload::Init - create head object entry for meta obj (src_obj_name + "." + upload_id)
[ Meta object stores all the parts upload info]

MultipartWriter::process - create all data/tail objects with obj_name same as
meta obj (so that they can all be identified & deleted during abort)

MultipartUpload::Abort - Just delete meta obj .. that will indirectly delete all the uploads
associated with that upload id / meta obj so far.

MultipartUpload::Complete - Create head object of the original object (if not exists).
Rename all data/tail object entries' obj name to orig object name and update metadata of the orig object.

Signed-off-by: Soumya Koduri <skoduri@redhat.com>

Merge pull request #44232 from Huber-ming/admin_fixlogs

radosgw-admin: fix some error logs

Merge pull request #43915 from qiuxinyidian/dev-rgw

rgw: when radosgw-admin stating user, add user exists judging

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

Merge pull request #43834 from Huber-ming/admin_bilog

radosgw-admin: supplement help documents with 'bilog autotrim'

doc/start: remove journal info from hardware recs

This PR removes mentions of journaling from the hardware
recommendations.

Journaling was a FileStore-related practice. BlueStore is
the default backend for Ceph OSDs and has been since
Luminous. The documentation should reflect that.

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

Merge pull request #44093 from melissa-kun-li/ssh-non-root-user

mgr/cephadm: support bootstrap with non-root ssh-user

Reviewed-by: Michael Fritch <mfritch@suse.com>
Reviewed-by: Sebastian Wagner <sewagner@redhat.com>

Merge PR #44162 into master

* refs/pull/44162/head:
mgr: only queue notify events that modules ask for
pybind/mgr: annotate which events modules consume
pybind/mgr: introduce NotifyType enum
mgr: stop issuing events that no modules consume

Reviewed-by: Sebastian Wagner <sewagner@redhat.com>

Merge PR #44196 into master

* refs/pull/44196/head:
mon/MgrStatMonitor: do not spam subscribers (mgr) with service_map

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

Merge PR #44207 into master

* refs/pull/44207/head:
mgr/ActivePyModule: avoid with_gil where possible
mgr/ActivePyModules: push without_gil_t down into blocks

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

Merge pull request #43542 from rzarzynski/wip-crimson-net-ms_learn_from_peer

crimson/net: add support for ms_learn_addr_from_peer.

Reviewed-by: Yingxin Cheng <yingxin.cheng@intel.com>
Reviewed-by: Chunmei Liu <chunmei.liu@intel.com>

Merge pull request #44104 from guits/guits-cephadm-skip-cv-restorecon

cephadm: pass `CEPH_VOLUME_SKIP_RESTORECON=yes`

Reviewed-by: Ken Dreyer <kdreyer@redhat.com>
Reviewed-by: Sebastian Wagner <sewagner@redhat.com>

Merge pull request #44239 from guits/guits-add-skip-needs-root

ceph-volume: make it possible to skip needs_root()

Merge pull request #44245 from tchaikov/wip-crimson-seastar

seastar: pick up change to fix FTBFS with old cryptopp

Reviewed-by: Samuel Just <sjust@redhat.com>
Reviewed-by: Yingxin Cheng <yingxin.cheng@intel.com>
Reviewed-by: Radoslaw Zarzynski <rzarzyns@redhat.com>

crimson/net: add support for ms_learn_addr_from_peer.

Signed-off-by: Radoslaw Zarzynski <rzarzyns@redhat.com>

Merge pull request #44179 from athanatos/sjust/wip-pin-race

crimson/os/seastore: initialize logical pins before exposing to cache

Reviewed-by: Yingxin Cheng <yingxin.cheng@intel.com>
Reviewed-by: Xuehan Xu <xxhdx1985126@gmail.com>

crimson/osd: s/seastar::fprint()/fmt::print()/

otherwise, we'd have warnings like:

./src/crimson/osd/main.cc:106:16: error: 'fprint<const std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > &>' is deprecated: use std::format_to() or fmt::print() [-Werror,-Wdeprecated-declarations]
      seastar::fprint(std::cerr, "already have key in keyring: %s\n", path);
               ^

Signed-off-by: Kefu Chai <tchaikov@gmail.com>

seastar: pick up change to fix FTBFS with old cryptopp

Signed-off-by: Kefu Chai <kefu@xsky.com>

Merge pull request #44186 from qiuxinyidian/rgw-de

rgw: add object null point judging when listing pubsub  topics

crimson/os/seastore: measure the number of conflicting transactions by srcs

Signed-off-by: Yingxin Cheng <yingxin.cheng@intel.com>

crimson/os/seastore: differentiate cleaner trim/reclaim transactions

Signed-off-by: Yingxin Cheng <yingxin.cheng@intel.com>

crimson/osd: fix heartbeat front and back blank ip

when ceph.conf not set public ip & cluster ip, heartbeat will get blank ip address. when osd::_send_boot , classic osd will check if heartbeat front and back addrs are blank ip, if they are blank ip, will use public ip which is learned from mon to set into them. So implement them in crimson osd.

Signed-off-by: chunmei-liu <chunmei.liu@intel.com>

Merge pull request #44147 from rzarzynski/wip-crimson-new-seastar

crimson: bump up Seastar to recent master and fix FTBFS

Reviewed-by: Samuel Just <sjust@redhat.com>
Reviewed-by: Chunmei Liu <chunmei.liu@intel.com>

Merge pull request #44095 from Matan-B/wip-matanb-local-workunits

doc/dev: Running workunits locally

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

Merge pull request #43940 from TRYTOBE8TME/wip-rgw-empty-config

src/rgw: Empty configuration support

Merge pull request #43665 from zenomri/wip-omri-multipart-trace

rgw/tracer: Multipart upload trace

Merge pull request #44156 from rzarzynski/wip-crimson-fix-process_op-sequencing

crimson/osd: fix sequencing issues in ClientRequest::process_op.

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

Merge pull request #44223 from rzarzynski/wip-crimson-fix-pullinfo-on-push

crimson/osd: don't assume a pull must happen if there is no push.

Reviewed-by: Samuel Just <sjust@redhat.com>
Reviewed-by: Chunmei Liu <chunmei.liu@intel.com>

Merge pull request #44224 from rzarzynski/wip-crimson-clean-msghs

crimson/osd: clean the recovery message-related header inclusion.

Reviewed-by: Chunmei Liu <chunmei.liu@intel.com>
Reviewed-by: Samuel Just <sjust@redhat.com>

ceph-volume: fix error 'KeyError' with inventory

The tag ceph.cluster_name is always set at the end.
The only way it could be absent was if the osd prepare
has been interrupted between [1] and [2].

[1] https://github.com/ceph/ceph/blob/v14.2.11/src/ceph-volume/ceph_volume/devices/lvm/strategies/bluestore.py#L355-L387
[2] https://github.com/ceph/ceph/blob/v14.2.11/src/ceph-volume/ceph_volume/devices/lvm/prepare.py

Although the code received tremendous changes meantime
and this error shouldn't show up again, we need to handle
the case where this tag wouldn't have been set.

Fixes: https://tracker.ceph.com/issues/44356
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>

Merge pull request #44184 from rzarzynski/wip-crimson-internal_client_request-fix-hobj

crimson/osd: fix assertion failure in InternalClientRequest.

Reviewed-by: Samuel Just <sjust@redhat.com>
Reviewed-by: Chunmei Liu <chunmei.liu@intel.com>

ceph-volume: fix tags dict output in `lvm list`

Default value for `--crush-device-class` is `None`.
When not passing this parameter, ceph-volume sets the
value "None" in the lv tags.
Therefore, ceph-volume will output that value with calling
`ceph-volume lvm list --format json`

For instance:
```
    "1": [
        {
            "devices": [
                "/dev/sdc"
            ],
            "lv_name": "osd-data-5a4a34f5-5733-4c69-b439-edb48e31a45f",
            "lv_path": "/dev/ceph-aeb16fc3-9ac2-4126-ab66-bf920d101ea4/osd-data-5a4a34f5-5733-4c69-b439-edb48e31a45f",
            "lv_size": "49.00g",
            "lv_tags": "ceph.block_device=/dev/ceph-aeb16fc3-9ac2-4126-ab66-bf920d101ea4/osd-data-5a4a34f5-5733-4c69-b439-edb48e31a45f,ceph.block_uuid=E9hZNU-80Zz-PiER-iWN3-jSIU-krEN-khwU3x,ceph.cephx_lockbox_secret=,ceph.cluster_fsid=40fe4af5-0408-444b-843c-0926d550d1f1,ceph.cluster_name=ceph,ceph.crush_device_class=None,ceph.encrypted=0,ceph.osd_fsid=39680838-19df-4e50-9bb6-46b093d5b52b,ceph.osd_id=1,ceph.type=block,ceph.vdo=0",
            "lv_uuid": "E9hZNU-80Zz-PiER-iWN3-jSIU-krEN-khwU3x",
            "name": "osd-data-5a4a34f5-5733-4c69-b439-edb48e31a45f",
            "path": "/dev/ceph-aeb16fc3-9ac2-4126-ab66-bf920d101ea4/osd-data-5a4a34f5-5733-4c69-b439-edb48e31a45f",
            "tags": {
                "ceph.block_device": "/dev/ceph-aeb16fc3-9ac2-4126-ab66-bf920d101ea4/osd-data-5a4a34f5-5733-4c69-b439-edb48e31a45f",
                "ceph.block_uuid": "E9hZNU-80Zz-PiER-iWN3-jSIU-krEN-khwU3x",
                "ceph.cephx_lockbox_secret": "",
                "ceph.cluster_fsid": "40fe4af5-0408-444b-843c-0926d550d1f1",
                "ceph.cluster_name": "ceph",
                "ceph.crush_device_class": "None",
```

ceph-volume should print `"ceph.crush_device_class": "",` instead of `"ceph.crush_device_class": "None",`

Fixes: https://tracker.ceph.com/issues/53425
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>

rgw: deleting objects inline in case bypass_gc is specified
for bucket remove command.

fixes: https://tracker.ceph.com/issues/53512

Signed-off-by: Pritha Srivastava <prsrivas@redhat.com>

ceph-volume: make it possible to skip needs_root()

Add the possibility to skip the `needs_root()` decorator.
See linked tracker for details.

Fixes: https://tracker.ceph.com/issues/53511
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>

Merge pull request #44145 from rhcs-dashboard/fix-frontend-vulnerabilities

mgr/dashboard: fix frontend deps' vulnerabilities

Reviewed-by: Waad Alkhoury <walkhour@redhat.com>
Reviewed-by: Pere Diaz Bou <pdiazbou@redhat.com>

crimson/osd: fix sequencing issues in ClientRequest::process_op.

The following crash has been observed in one of the runs at Sepia:

```
ceph-osd: /home/jenkins-build/build/workspace/ceph-dev-new-build/ARCH/x86_64/AVAILABLE_ARCH/x86_64/AVAILABLE_DIST/centos8/DIST/centos8/MACHINE_SIZE/gigantic/release/17.0.0-8898-ge57ad63c/rpm/el8/BUILD/ceph-17.0.0-8898-ge57ad63c/src/crimson/osd/osd_operation_sequencer.h:123: void crimson::osd::OpSequencer::finish_op_in_order(crimson::osd::ClientRequest&): Assertion `op.get_id() > last_completed_id' failed.
Aborting on shard 0.
Backtrace:
Reactor stalled for 1807 ms on shard 0. Backtrace: 0xb14ab 0x46e57428 0x46bc450d 0x46be03bd 0x46be0782 0x46be0946 0x46be0bf6 0x12b1f 0x137341 0x3fdd6a92 0x3fddccdb 0x3fdde1ee 0x3fdde8b3 0x3fdd3f2b 0x3fdd4442 0x3fdd4c3a 0x12b1f 0x3737e 0x21db4 0x21c88 0x2fa75 0x3b769527 0x3b8418af 0x3b8423cb 0x3b842ce0 0x3b84383d 0x3a116220 0x3a143f31 0x3a144bcd 0x46b96271 0x46bde51a 0x46d6891b 0x46d6a8f0 0x4681a7d2 0x4681f03b 0x39fd50f2 0x23492 0x39b7a7dd
 0# gsignal in /lib64/libc.so.6
 1# abort in /lib64/libc.so.6
 2# 0x00007FB9FB946C89 in /lib64/libc.so.6
 3# 0x00007FB9FB954A76 in /lib64/libc.so.6
 4# 0x00005595E98E6528 in ceph-osd
 5# 0x00005595E99BE8B0 in ceph-osd
 6# 0x00005595E99BF3CC in ceph-osd
 7# 0x00005595E99BFCE1 in ceph-osd
 8# 0x00005595E99C083E in ceph-osd
 9# 0x00005595E8293221 in ceph-osd
10# 0x00005595E82C0F32 in ceph-osd
11# 0x00005595E82C1BCE in ceph-osd
12# 0x00005595F4D13272 in ceph-osd
13# 0x00005595F4D5B51B in ceph-osd
14# 0x00005595F4EE591C in ceph-osd
15# 0x00005595F4EE78F1 in ceph-osd
16# 0x00005595F49977D3 in ceph-osd
17# 0x00005595F499C03C in ceph-osd
18# main in ceph-osd
19# __libc_start_main in /lib64/libc.so.6
20# _start in ceph-osd
```

The sequence of events provides at least two clues:
  - the op no. 32 finished before the op no. 29 which was waiting
    for `ObjectContext`,
  - the op no. 29 was a short-living one -- it wasn't waiting even
    on `obc`.

```
rzarzynski@teuthology:/home/teuthworker/archive/rzarzynski-2021-11-22_22:01:32-rados-master-distro-basic-smithi$ less ./6520106/remote/smithi115/log/ceph-osd.3.log.gz
...
DEBUG 2021-11-22 22:32:24,531 [shard 0] osd - client_request(id=29, detail=m=[osd_op(client.4371.0:36 4.d 4.f0fb5e1d (undecoded) ondisk+retry+read+rwordered+known_if_redirected+supports_pool_eio e23) v8]): start
DEBUG 2021-11-22 22:32:24,531 [shard 0] osd - client_request(id=29, detail=m=[osd_op(client.4371.0:36 4.d 4.f0fb5e1d (undecoded) ondisk+retry+read+rwordered+known_if_redirected+supports_pool_eio e23) v8]): in repeat
...
DEBUG 2021-11-22 22:32:24,546 [shard 0] osd - client_request(id=29, detail=m=[osd_op(client.4371.0:36 4.d 4.f0fb5e1d (undecoded) ondisk+retry+read+rwordered+known_if_redirected+supports_pool_eio e23) v8]) same_interval_since: 21
DEBUG 2021-11-22 22:32:24,546 [shard 0] osd - OpSequencer::start_op: op=29, last_started=27, last_unblocked=27, last_completed=27
...
DEBUG 2021-11-22 22:32:24,621 [shard 0] osd - client_request(id=32, detail=m=[osd_op(client.4371.0:49 4.d 4.81addbad (undecoded) ondisk+retry+write+known_if_redirected+supports_pool_eio e23) v8]): start
DEBUG 2021-11-22 22:32:24,621 [shard 0] osd - client_request(id=32, detail=m=[osd_op(client.4371.0:49 4.d 4.81addbad (undecoded) ondisk+retry+write+known_if_redirected+supports_pool_eio e23) v8]): in repeat
...
DEBUG 2021-11-22 22:32:24,626 [shard 0] osd - client_request(id=32, detail=m=[osd_op(client.4371.0:49 4.d 4.81addbad (undecoded) ondisk+retry+write+known_if_redirected+supports_pool_eio e23) v8]) same_interval_s
ince: 21
DEBUG 2021-11-22 22:32:24,626 [shard 0] osd - OpSequencer::start_op: op=32, last_started=29, last_unblocked=29, last_completed=27
<note that op 32 is very short living>
DEBUG 2021-11-22 22:32:24,669 [shard 0] osd - OpSequencer::finish_op_in_order: op=32, last_started=32, last_unblocked=32, last_completed=27
...
DEBUG 2021-11-22 22:32:24,671 [shard 0] osd - client_request(id=32, detail=m=[osd_op(client.4371.0:49 4.d 4:b5dbb581:::smithi11538976-13:head {write 601684~619341 in=619341b, stat} snapc 0={} RETRY=1 ondisk+retry+write+known_if_redirected+supports_pool_eio e23) v8]): destroying
...
DEBUG 2021-11-22 22:32:24,722 [shard 0] osd - client_request(id=29, detail=m=[osd_op(client.4371.0:36 4.d 4:b87adf0f:::smithi11538976-9:head {read 0~1} snapc 0={} RETRY=1 ondisk+retry+read+rwordered+known_if_redirected+supports_pool_eio e23) v8]): got obc lock
...
INFO  2021-11-22 22:32:24,723 [shard 0] osd - client_request(id=29, detail=m=[osd_op(client.4371.0:36 4.d 4:b87adf0f:::smithi11538976-9:head {read 0~1} snapc 0={} RETRY=1 ondisk+retry+read+rwordered+known_if_redirected+supports_pool_eio e23) v8]) obc.get()=0x6190000d5780
...
DEBUG 2021-11-22 22:32:24,753 [shard 0] osd - OpSequencer::finish_op_in_order: op=29, last_started=32, last_unblocked=32, last_completed=32
ceph-osd: /home/jenkins-build/build/workspace/ceph-dev-new-build/ARCH/x86_64/AVAILABLE_ARCH/x86_64/AVAILABLE_DIST/centos8/DIST/centos8/MACHINE_SIZE/gigantic/release/17.0.0-8898-ge57ad63c/rpm/el8/BUILD/ceph-17.0.0-8898-ge57ad63c/src/crimson/osd/osd_operation_sequencer.h:123: void crimson::osd::OpSequencer::finish_op_in_order(crimson::osd::ClientRequest&): Assertion `op.get_id() > last_completed_id' failed.
Aborting on shard 0.
```

This could be explained in a scenario where:
  - op no. 29 skipped stages of the execution pipeline while
  - it wrongly informed `OpSequencer` the execution was in-order.

Static analysis shows there are multiple problems of this genre
in the `ClientRequest::process_op()` and its callees with the most
recently merged one being the path for `PG::already_complete()`.

Signed-off-by: Radoslaw Zarzynski <rzarzyns@redhat.com>

crimson/os/seastore/lba_manager: initialize lba node pins using get_extent

Signed-off-by: Samuel Just <sjust@redhat.com>

crimson/os/seastore: initialize logical pins before exposing to cache

Otherwise, another task may get a reference to the extent before
we've set the pin.

Fixes: https://tracker.ceph.com/issues/53267
Signed-off-by: Samuel Just <sjust@redhat.com>

Merge pull request #44231 from xxhdx1985126/wip-cpu-profile

crimson/os/seastore: fix compiler error for gcc > 9 and clang13

Reviewed-by: Yingxin Cheng <yingxin.cheng@intel.com>
Reviewed-by: Samuel Just <sjust@redhat.com>

radosgw-admin: fix some error logs

Signed-off-by: Huber-ming <zhangsm01@inspur.com>

crimson/os/seastore: fix compiler error for gcc > 9 and clang13

Signed-off-by: Xuehan Xu <xxhdx1985126@gmail.com>

osd/OSDMapMapping: fix spurious threadpool timeout errors

We were passing a grace of zero seconds to our temporary work queue, which
led to the HeartbeatMap issuing cpu_tp timeout errors to the log.  By using
a non-zero grace period we can avoid these.  Use the same default grace
we use for the workqueue itself when it goes to sleep.

Fixes: https://tracker.ceph.com/issues/53506
Signed-off-by: Sage Weil <sage@newdream.net>