Merge pull request #47002 from rzarzynski/wip-mgr-fix-daemon-server-race

mgr: fix a race condition in DaemonServer::handle_report()

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

Merge pull request #46933 from clwluvw/s3-website-empty-bucket

rgw: s3website check for bucket before retargeting

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

Merge pull request #47009 from adamemerson/wip-chatty-bucket-sync-run

rgw: User friendly progress report on bucket sync run

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

Merge pull request #47163 from rhcs-dashboard/get-storage-class-metrics

mgr/dashboard:Get different storage class metrics in Prometheus dashboard

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

Merge pull request #46993 from rishabh-d-dave/qa-caps-helper-multiple-errmsgs

qa/cephfs: allow checking for multiple error messages while...

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

Merge pull request #47185 from cyx1231st/wip-seastore-backref-block-size

crimson/os/seastore/backref: decouple SegmentManagerGroup

Reviewed-by: Myoungwon Oh <myoungwon.oh@samsung.com>
Reviewed-by: Xuehan Xu <xxhdx1985126@gmail.com>

Merge pull request #47183 from cyx1231st/wip-seastore-cleaner-journal-head

crimson/os/seastore/async_cleaner: re-introduce journal_header member

Reviewed-by: Myoungwon Oh <myoungwon.oh@samsung.com>

Merge pull request #47125 from xxhdx1985126/wip-56535

crimson/os/seastore/async_cleaner: avoid trimming backrefs when reclaiming space

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

qa/cephfs: allow checking for multiple error messages while...

while negative testing.

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

Merge pull request #47013 from rishabh-d-dave/qa-caps-method-authorize

qa/cephfs: upgrade method authorize in class Filesystem

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

Merge pull request #46991 from rishabh-d-dave/qa-caps-helper-rm-path

qa/cephfs: delete path from cmd args after use

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

Merge pull request #46998 from ceph/dependabot-npm_and_yarn-src-pybind-mgr-dashboard-frontend-moment-2.29.4

mgr/dashboard: bump moment from 2.29.3 to 2.29.4 in /src/pybind/mgr/dashboard/frontend

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

Merge pull request #42335 from rishabh-d-dave/qa-get-keyring

qa: add a method to get keyring for a client_id

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

mgr/dashboard:Get "Different Storage Class" metrics in Prometheus dashboard

Get metrics of the different "HDDRule" and "MixedUse" classes of the "Raw Storage" for their ceph VMs. So that Prometheus can scrape the data and display it to them in grafana

Fixes: https://tracker.ceph.com/issues/56625
Signed-off-by: Aashish Sharma <aasharma@redhat.com>

crimson/os/seastore/backref: decouple SegmentManagerGroup

Backref tree is not specific to segments.

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

crimson/os/seastore: drop unused code

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

Merge pull request #47137 from rhcs-dashboard/cephadm-e2e-more-fixes

mgr/dashboard: fix dashboard cephadm e2e errors

Reviewed-by: Pedro González <pegonzal@redhat.com>
Reviewed-by: Ernesto Puerta <epuertat@redhat.com>

crimson/os/seastore/async_cleaner: re-introduce journal_header member

RBM won't have segments_info_t, it has to maintain the journal_header by
itself.

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

Merge pull request #47173 from alimaredia/wip-rgw-s3tests-cloudtransitions-iam-section-add

qa: add iam section to cloud transitions config

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

crimson/os/seastore/backref_manager: rename some methods for the sake of readability

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

crimson/os/seastore/async_cleaner: don't trim backref when reclaiming space

Since the current backref cache doesn't invalidate duplicated backrefs any more
and backrefs get trimmed in the exact order of journal seqs, there's no need to trim
backrefs when reclaiming space

Fixes: https://tracker.ceph.com/issues/56535
Signed-off-by: Xuehan Xu <xxhdx1985126@gmail.com>

Merge pull request #47140 from cyx1231st/wip-seastore-split-fix

crimson/os/seastore: do not assign gen/hint when split a clean extent

Reviewed-by: Samuel Just <sjust@redhat.com>
Reviewed-by: Zhang Song <zhangsong325@gmail.com>

qa: add iam section to cloud transitions config

Signed-off-by: Ali Maredia <amaredia@redhat.com>

Merge pull request #46736 from baergj/fix-log-buffer-resize

log: Make log_max_recent have an effect again

Reviewed-by: Kefu Chai <tchaikov@gmail.com>
Reviewed-by: Casey Bodley <cbodley@redhat.com>

Merge pull request #46666 from MrFreezeex/batch-blockdb-fix

ceph-volume: fix fast device alloc size on mulitple device

Merge pull request #46928 from linuxbox2/wip-rgwlc-azone

rgwlc: permit lifecycle to reduce data conditionally in archive zone

Merge pull request #47019 from badone/wip-get_or_fail-debug-louder

msg: Log at higher level when Throttle::get_or_fail() fails

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

Merge pull request #46866 from myoungwon/wip-fix-53294

osd: return ENOENT if pool information is invalid during tier-flush

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

Merge pull request #47116 from chrisphoffman/wip-rbd-56549

librbd: bail from schedule_request_lock() if already lock owner

Reviewed-by: Mykola Golub <mgolub@suse.com>
Reviewed-by: Ilya Dryomov <idryomov@gmail.com>

Merge pull request #47139 from rishabh-d-dave/cephfs-top-man

doc/man/cephfs-top.rst: add missing options: --delay, --conffile

Reviewed-by: Anthony D'Atri <anthony.datri@gmail.com>
Reviewed-by: Jos Collin <jcollin@redhat.com>
Reviewed-by: Dhairya Parmar <dparmar@redhat.com>

Merge pull request #46889 from rhcs-dashboard/osd-followup-iops

mgr/dashboard: do not recommend throughput for ssd's only cluster

Reviewed-by: Ernesto Puerta <epuertat@redhat.com>
Reviewed-by: Pere Diaz Bou <pdiazbou@redhat.com>
Reviewed-by: sunilangadi2 <NOT@FOUND>

Merge pull request #47135 from rhcs-dashboard/doc-default-main

mgr/dashboard: change doc service default release from master to main

Merge pull request #47034 from tchaikov/wip-mon-signness-cleanup

mon: make paxos_size() unsigned

Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
Reviewed-by: Radoslaw Zarzynski <rzarzyns@redhat.com>
Reviewed-by: Kamoltat Sirivadhna <ksirivad@redhat.com>
Reviewed-by: Laura Flores <lflores@redhat.com>

Merge pull request #47089 from athanatos/sjust/wip-pg-shard-manager

crimson: introduce pg_shard_manager, begin to separate osd-wide from core-local state

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

doc/man/cephfs-top.rst: add missing options: --delay, --conffile

add missing options: --delay, --conffile

Signed-off-by: wangxinyu <wangxinyu@inspur.com>
Signed-off-by: Rishabh Dave <ridave@redhat.com>

Merge pull request #47152 from tchaikov/wip-crimson-close-later

crimson/net: postpone the close() using yield()

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

Merge pull request #46493 from adk3798/tuned-profiles

mgr/cephadm: support for os tuning profiles

Reviewed-by: Anthony D'Atri <anthonyeleven@users.noreply.github.com>
Reviewed-by: John Mulligan <jmulligan@redhat.com>
Reviewed-by: Redouane Kachach <rkachach@redhat.com>

Merge pull request #46924 from hookak/loki-support

mgr/cephadm: fix the loki address in grafana, promtail configuration file

Reviewed-by: Aashish Sharma <aasharma@redhat.com>
Reviewed-by: Adam King <adking@redhat.com>

Merge pull request #46620 from rzarzynski/wip-tools-cot-force-pg-import

tools: COT ignores fsid mismatch when importing PG with --force

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

Merge pull request #47044 from alimaredia/wip-rgw-suite-s3test-refactor

add s3tests-brach.yaml for rgw teuthology suites that run s3tests

Reviewed-by: Ali Maredia <amaredia@redhat.com>

Merge pull request #46615 from selvakumaar5496/main

rgw: fix api response in case of get and delete object tagging apis

Reviewed-by: Ali Maredia <amaredia@redhat.com>

crimson/net: postpone the close() using yield()

otherwise we'd erase an element in a container when we are still
iterating through it.

Fixes: https://tracker.ceph.com/issues/56589
Signed-off-by: Kefu Chai <tchaikov@gmail.com>

Merge pull request #47136 from idryomov/wip-48038

qa/suites/rbd: disable workunit timeout for dynamic_features_no_cache

Reviewed-by: Deepika Upadhyay <dupadhya@redhat.com>

librbd: bail from schedule_request_lock() if already lock owner

Race condition may be hit if there are multiple pending locks for the
same image and pending callbacks. Abort exclusive lock process if
already exclusive lock owner.

Fixes: https://tracker.ceph.com/issues/56549
Signed-off-by: Christopher Hoffman <choffman@redhat.com>

Merge pull request #46911 from ifed01/wip-ifed-fix-mempool-cache-other

os/bluestore: fix AU accounting in bluestore_cache_other mempool.

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

Merge pull request #47106 from idryomov/wip-56561

rbd: don't default empty pool name unless namespace is specified

Reviewed-by: Christopher Hoffman <choffman@redhat.com>
Reviewed-by: Mykola Golub <mgolub@suse.com>

Merge pull request #46741 from Matan-B/wip-matanb-docker-debug

script: CentOS 8 EOL, use archived mirror in ceph-debug-docker.sh

Reviewed-by: Nitzan Mordechai nmordech@redhat.com
Reviewed-by: Radoslaw Zarzynski <rzarzyns@redhat.com>

Merge pull request #46375 from ceph/cv-loop-devs

ceph-volume: Optionally consume loop devices

crimson/os/seastore: do not assign gen/hint when split a clean extent

Gen/hint is only required when need to write the extent to disk.

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

mgr/dashboard: do not recommend throughput for ssd's only cluster

This is just a bug fix where we recommend the throughput option even if
there are only ssd's are present in the cluster.

Fixes: https://tracker.ceph.com/issues/56413
Signed-off-by: Nizamudeen A <nia@redhat.com>

crimson/os/seastore: introduce PLACEMENT_HINT_NULL

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

Merge pull request #47113 from zhscn/fix-split-test

crimson/os/seastore: fix bugs in test_map_existing_extent_concurrent

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

rgwlc: return std::string_view from sal::Zone::get_tier_type()

Valid values are all small strings, often static.

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

rgwlc:  count LCFilter flags towards multi-condition

Found by Soumya Koduri <skoduri@redhat.com>

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

rgwlc: return real default zone type from sal_dbstore and sal_motr

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

rgwlc:  update LCFilter::dump_xml(...) to add flags/ArchiveZone

Also updates the location of Prefix, which is supposed to *generate*
as <Filter><Prefix/></Filter>, regardless of how we parsed it.

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

rgwlc: activate lifecycle processing on non-master zones

The basic idea of this change is the same as the proposal by
Ilsoo Byun <ilsoobyun@linecorp.com>, but some details have changed.

The main differences are to use the existing
RGWLC::set(remove)_bucket_config methods, and to use the
RGWBucketInstanceMetadataHandler infrastructue to dispatch
the corresponding calls.  Thank you!

Fixes: https://tracker.ceph.com/issues/44268
Related PR: #33524

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

rgwlc:  zone-conditional lifecycle processing

Lifecycle rules with the ArchiveZone flag must execute on archive zones,
but must not execute on others.

Fixes: https://tracker.ceph.com/issues/56440
Signed-off-by: Matt Benjamin <mbenjamin@redhat.com>

rgwlc: represent lc filter flags as XML elements

Suggested by Casey in review, this makes the XML prettier.

also: fix filter parsing, remove unused code

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

rgwlc: introduce lifecycle config flags extension

rgwlc: add uint32_t flags bitmap to LCFilter

This is intended to support a concise set of extensions to S3
LifecycleConfiguration, initially, just a flag that indicates a
rule is intended for execution on RGW ArchiveZone.

rgwlc: add machinery to define and recognize LCFilter flags

Add a concept of filter flags to lifecycle filter rules, an RGW
extension.  The initial purpose of flags is to permit marking
specific lifecycle rules as specific to an RGW archive zone, but
other flags could be added in future.

rgwlc: add new unittest_rgw_lc to run internal checks, add a few
valid and invalid lifecycle configuration xml  parses for now.

Fixes: https://tracker.ceph.com/issues/53361
Signed-off-by: Matt Benjamin <mbenjamin@redhat.com>

mgr/dashboard: fix dashboard cephadm e2e errors

https://tracker.ceph.com/issues/56079
Signed-off-by: Nizamudeen A <nia@redhat.com>

qa/suites/rbd: disable workunit timeout for dynamic_features_no_cache

The I/O workload in this test is xfstests (qa/run_xfstests_qemu.sh)
which isn't subjected to any timeout other than global max_job_time
limit in any other subsuite (e.g. qemu/workloads/qemu_xfstests.yaml).
But here, there is a parallel "op" workload defined as a workunit.
The workunit task has a default timeout of 3 hours which is effectively
imposed on the entire job.  In the "rbd cache = false" configuration,
it's sometimes exceeded.

Fixes: https://tracker.ceph.com/issues/48038
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

mgr/dashboard: change doc service default release from master to main

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

rgw: Print syncing objects during `bucket sync run`

When --extra-info is provided to radosgw-admin.

This is a bit minimal but they weren't too clear in precisely what
information they wanted, and this should give a definite indication of
forward progress.

Signed-off-by: Adam C. Emerson <aemerson@redhat.com>

rgw: Print user friendly progress on `bucket sync init`

When `--extra-info` is provided

Signed-off-by: Adam C. Emerson <aemerson@redhat.com>

rgw: RGWBucketPipeSyncStatusManger takes optional ostream

`radosgw-admin` provides it when the `--extra-info` option is provided.

It gets stashed in the SyncEnv for later use.

Signed-off-by: Adam C. Emerson <aemerson@redhat.com>

rgw: RGWBucketPipeSyncStatusManager gets zone name on construction

So we can print it in the user friendly messages to come.

Signed-off-by: Adam C. Emerson <aemerson@redhat.com>

rgw: Fix interface of RGWSI_Zone::find_zone

Since nullable pointers are nullable, it makes no sense to return a
bool and take a nullable pointer out parameter.

Just return the pointer.

Signed-off-by: Adam C. Emerson <aemerson@redhat.com>

Merge pull request #47025 from adamemerson/wip-55765

rgw: Guard against malformed bucket URLs

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

Merge pull request #46897 from rkachach/fix_issue_55808

mgr/cephadm: check for events key before accessing it

Reviewed-by: Adam King <adking@redhat.com>
Reviewed-by: John Mulligan <jmulligan@redhat.com>

doc/cephadm: os tuning profile documentation

Signed-off-by: Adam King <adking@redhat.com>

mgr/cephadm: unit tests for tuned os profiles

Signed-off-by: Adam King <adking@redhat.com>

mgr/cephadm: support for os tuning profiles

Fixes: https://tracker.ceph.com/issues/55819
Signed-off-by: Adam King <adking@redhat.com>

Merge pull request #47011 from s0nea/wip-prevent-alert-redirects

mgr/dashboard: prevent alert redirect

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

Merge pull request #44217 from CongMinYin/fix-pwl-recovery-test

qa/suites/rbd/pwl-cache: ensure recovery is actually tested

Reviewed-by: Ilya Dryomov <idryomov@gmail.com>
Reviewed-by: Deepika Upadhyay <dupadhya@redhat.com>

crimson/os/seastore/segment_cleaner: trim backrefs first

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

rbd: drop unused default_empty_pool_name argument

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

rbd: don't default empty pool name unless namespace is specified

Commit 96f05a7956b3 ("rbd: delay determination of default pool name")
broke "rbd perf image iostat" and "rbd perf image iotop" GLOBAL_POOL_KEY
support (the ability to blend all rbd pools together into a single
view).

Fixes: https://tracker.ceph.com/issues/56561
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

qa/tasks: rename persistent write log cache trash task

It doesn't really thrash anything, just repeatedly restarts the
workload on top of a dirty cache file.  rbd_pwl_cache_recovery is
more on point and gets covered by existing CODEOWNERS.

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

mon: make paxos_size() unsigned

* make paxos_size() unsigned, as paxos_size() returns the size of
  MonMap::mon_info, so it should be always a non-negative value,
  and more importantly, it represents a size.
* change the type of MonMap::removed_ranks from std::set<int>
  to std::set<unsigned>. for two reasons:
  - removed_ranks only tracks the rank which is greater or equal to 0
  - helps to silence the warnings listed below.
  MonMap::removed_ranks is persisted using encode()/decode(), but this
  change is backward compatible, as we use the raw encoder to encode
  signed and unsigned integers, the difference between the encoding
  schema between them only matters when MSB in the number is used,
  but this is not likely happen, as we neither have a negative
  rank in removed_ranks, no have a rank greater than `(unsigned)-1`,
  i.e., 0xffffffff.

this change partially reverts f75dfbc055ccf4e43b817ed5aa52898ff680e19e

to address the compiling warnings like:

/home/kefu/dev/ceph/src/mon/ElectionLogic.cc: In member function ‘void ElectionLogic::end_election_period()’:
/home/kefu/dev/ceph/src/mon/ElectionLogic.cc:173:23: error: comparison of integer expressions of different signedness: ‘std::set<int>::size_type’ {aka ‘long unsigned int’} and ‘int’ [-Werror=sign-compare]
  173 |       acked_me.size() > (elector->paxos_size() / 2)) {
      |       ~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~
/home/kefu/dev/ceph/src/mon/ElectionLogic.cc: In member function ‘void ElectionLogic::propose_connectivity_handler(int, epoch_t, const ConnectionTracker*)’:
/home/kefu/dev/ceph/src/mon/ElectionLogic.cc:338:28: error: comparison of integer expressions of different signedness: ‘unsigned int’ and ‘int’ [-Werror=sign-compare]
  338 |     for (unsigned i = 0; i < elector->paxos_size(); ++i) {
      |                          ~~^~~~~~~~~~~~~~~~~~~~~~~
/home/kefu/dev/ceph/src/mon/ElectionLogic.cc: In member function ‘void ElectionLogic::receive_ack(int, epoch_t)’:
/home/kefu/dev/ceph/src/mon/ElectionLogic.cc:469:25: error: comparison of integer expressions of different signedness: ‘std::set<int>::size_type’ {aka ‘long unsigned int’} and ‘int’ [-Werror=sign-compare]
  469 |     if (acked_me.size() == elector->paxos_size()) {
      |         ~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~
cc1plus: all warnings being treated as errors
make[3]: *** [src/mon/CMakeFiles/mon.dir/build.make:328: src/mon/CMakeFiles/mon.dir/ElectionLogic.cc.o] Error 1
make[3]: *** Waiting for unfinished jobs....
make[3]: Leaving directory '/home/kefu/dev/ceph/build'
[ 48%] Built target libglobal_objs
/home/kefu/dev/ceph/src/mon/Elector.cc: In member function ‘void Elector::notify_rank_removed(int)’:
/home/kefu/dev/ceph/src/mon/Elector.cc:734:43: error: comparison of integer expressions of different signedness: ‘unsigned int’ and ‘int’ [-Werror=sign-compare]
  734 |     for (unsigned i = rank_removed + 1; i <= paxos_size() ; ++i) {
      |                                         ~~^~~~~~~~~~~~~~~

Fixes: https://tracker.ceph.com/issues/56581
Signed-off-by: Kefu Chai <tchaikov@gmail.com>

Merge pull request #47109 from zdover23/wip-doc-2022-07-15-start-hw-recs-cleanup-1

doc/start: update hardware recs

Reviewed-by: Anthony D'Atri

Merge pull request #46908 from mlausch/snapshot_key_conversion

osd/SnapMapper: fix legacy key conversion in snapmapper class

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

rgw: Guard against malformed bucket URLs

Misplaced colons can result in radosgw thinking is has a bucket URL
but with no bucket name, leading to a crash later on.

Fixes: https://tracker.ceph.com/issues/55765
Signed-off-by: Adam C. Emerson <aemerson@redhat.com>

rgw: Fix `rgw::sal::Bucket::empty` static method signatures

`unique_ptr` overload should take by reference.

Both should be const.

Signed-off-by: Adam C. Emerson <aemerson@redhat.com>

Merge pull request #47052 from neha-ojha/wip-cot-label

.github/labeler.yml: add core label to some tools

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

Merge pull request #47115 from ceph/fix-mib

ceph.spec.in: fix path for mib file and properly mark in %files

Reviewed-by: Kefu Chai <tchaikov@gmail.com>
Reviewed-by: Ilya Dryomov <idryomov@gmail.com>

Merge pull request #47035 from idryomov/wip-56516

rbd-mirror: remove bogus completed_non_primary_snapshots_exist check

Reviewed-by: Mykola Golub <mgolub@suse.com>
Reviewed-by: Deepika Upadhyay <dupadhya@redhat.com>

ceph.spec.in: fix path for mib file and properly mark in %files

Fixes typos introduced in https://github.com/ceph/ceph/pull/46918

Signed-off-by: Justin Caratzas <jcaratza@redhat.com>

osd: return ENOENT if pool information is invalid during tier-flush

During tier-flush, OSD sends reference increase message to target OSD.
At this point, sending message with invalid pool information (e.g., deleted pool)
causes unexpected behavior.

Therefore, this commit return ENOENT early before sending the message

fixes: https://tracker.ceph.com/issues/53294

Signed-off-by: Myoungwon Oh <myoungwon.oh@samsung.com>

Merge pull request #47028 from xxhdx1985126/wip-seastore-backref-cache-refactor

crimson/os/seastore: simplify backref cache

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

crimson/os/seastore: fix bugs in test_map_existing_extent_concurrent

Signed-off-by: Zhang Song <zhangsong325@gmail.com>

crimson/os/seastore: simplify backref cache

Currently, the following transaction exec sequence would lead to
loss of backref:

1. Trans `A` merge a alloc backref for extent `X`
2. Trans `B` add a release backref for extent `X` to backref cache,
   during which it finds an in-cache alloc backref for extent `X` and
   decide not to add the release backref to cache
3. Trans `A` commit

In the above sequece, the release backref for extent `X` is lost.

This is a regression introduced when we try to optimize the backref cache.

This commit fix the issue by caching inflight backrefs in a multiset,
alloc/release ops that happen on the same paddr are queued in the order of
their happening. When doing gc, all those backrefs are merged.

Fixes: https://tracker.ceph.com/issues/56519
Signed-off-by: Xuehan Xu <xxhdx1985126@gmail.com>

doc/start: update hardware recs

This PR picks up the parts of
https://github.com/ceph/ceph/pull/44466
that were not merged back in January, when that
pull request was raised.

Matters added here:
* improved organzation of matter
* emphasis of IOPs per core over cores per OSD

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

Merge pull request #46546 from mgfritch/vstart-stop-mds

src/stop.sh: stop existing ceph-mds daemons

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

Merge pull request #46014 from vrushch/rbd_form_fix

mgr/dashboard: rbd striping setting pre-population and pop-over

Reviewed-by: Pegonzal <NOT@FOUND>
Reviewed-by: Ernesto Puerta <epuertat@redhat.com>
Reviewed-by: Nizamudeen A <nia@redhat.com>
Reviewed-by: Pere Diaz Bou <pdiazbou@redhat.com>

Merge pull request #47066 from adk3798/osd-removal-docs-update

doc/cephadm: add note about OSDs being recreated to OSD removal section

Reviewed-by: Anthony D'Atri <anthonyeleven@users.noreply.github.com>
Reviewed-by: Redouane Kachach <rkachach@redhat.com>

Merge pull request #47062 from guits/update-latest-stable-default

cephadm: update LATEST_STABLE_RELEASE

Reviewed-by: Adam King <adking@redhat.com>

Merge pull request #46260 from CongMinYin/wip-enable-ndctl

cmake: enable ndctl when building PMDK for WITH_BLUESTORE_PMEM

Reviewed-by: Kefu Chai <tchaikov@gmail.com>
Reviewed-by: Ilya Dryomov <idryomov@gmail.com>

Merge pull request #46644 from rhcs-dashboard/rbd-list-pagination

mgr/dashboard: rbd image pagination

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

Merge pull request #45981 from rhcs-dashboard/box-no-more-ultron

cephadm/box: Rootless podman box implementation

Reviewed-by: Pegonzal <NOT@FOUND>
Reviewed-by: anthonyeleven <NOT@FOUND>
Reviewed-by: melissa-kun-li <NOT@FOUND>