doc: document rank option for journal reset

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

Merge pull request #31180 from liewegas/revert-bluestore-pinned-list

os/bluestore: revert cache pinned list

Reviewed-by: Mark Nelson <mnelson@redhat.com>

Merge pull request #31071 from sebastian-philipp/python-common-mypy

python-common: Add mypy testing

Reviewed-by: Jan Fajerski <jfajerski@suse.com>

Merge pull request #30995 from wjwithagen/wjw-wip-volume-zfs-inventory

ceph-volume-zfs: add the inventory command

Merge pull request #31014 from tchaikov/wip-38378

mgr/dashboard: accept expected exception when SSL handshaking

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

Merge pull request #31092 from liewegas/wip-revert-42026-test

Revert "test: librados startup/shutdown racer test"

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

Merge pull request #31153 from ceph/wip-restful-node-items

restful: Query nodes_by_id for items

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

Merge pull request #31176 from tchaikov/wip/mgr/diskprediction_cloud/cleanup

mgr/diskprediction_cloud: refactor timeout() decorator

Reviewed-by: Rick Chen <rick.chen@prophetstor.com>
Reviewed-by: Ernesto Puerta <epuertat@redhat.com>

Merge pull request #31175 from tchaikov/wip/mon/OSDMonitor/cleanup

mon/OSDMonitor: use initializer_list<> for {si,iec}_options

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

Merge pull request #30279 from runsisi/wip-fix-strerror_r

common: fix compat of strerror_r

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

Merge pull request #29836 from majianpeng/fix-async-msgr-perfcounter-bug

msg/async: Don't dec(msgr_active_connections) if conn still in accept…

Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
Reviewed-by: Ricardo Dias <rdias@suse.com>

Merge pull request #31156 from cyx1231st/wip-seastar-msgr-testpeer

test/crimson: implement a remote async TestPeer for crimson msgr tests

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

test/crimson: add nonstop option in async TestPeer

If set, TestPeer won't shutdown if all tests are done, and will wait for
another round of tests.

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

test/crimson: add delay to markdown_peer()

Let test sleep awhile for peer markdown propagated.

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

test/crimson: async msgr TestPeer for v2 failover tests

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

test/crimson: allow v2 failover tests to connect to remote TestPeer

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

test/crimson: extract msgr test commands to header

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

Merge pull request #30610 from ivancich/wip-no-schedule-reshard-during-reshard

rgw: prevent bucket reshard scheduling if bucket is resharding

Reviewed-by: Matt Benjamin <mbenjamin@redhat.com>
Reviewed-by: Casey Bodley <cbodley@redhat.com>
Reviewed-by: Abhishek Lekshmanan <abhishek@suse.com>

Merge pull request #30795 from ivancich/wip-dynamic-reshard-adjustments

rgw: adjust allowable bucket index shard counts for dynamic resharding

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

Merge pull request #30845 from croit/rgw-bucket-stats-add-num-shards

rgw: add num_shards to radosgw-admin bucket stats

Reviewed-by: Casey Bodley <cbodley@redhat.com>
Reviewed-by: Abhishek Lekshmanan <abhishek@suse.com>
Reviewed-by: J. Eric Ivancich <ivancich@redhat.com>

Merge pull request #30956 from joke-lee/rgw-storage-class-post-obj

rgw: set correct storage class for post object upload

Reviewed-by: Abhishek Lekshmanan <abhishek@suse.com>

Merge pull request #30970 from theanalyst/rgw/list-common-prefix

rgw: url encode common prefixes for List Objects response

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

Merge pull request #31179 from liewegas/cleanup-exec-bit

os/bluestore,comon,erasure-code: chmod -x source files

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

Merge pull request #31171 from liewegas/bug-42496

qa/tasks/cbt: run stop-all.sh while shutting down

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

Revert "Merge pull request #30964 from markhpc/wip-bs-cache-trim-pinned"

This reverts commit 304c37f52165363cad4aaf31386d62ea5a573425, reversing
changes made to 7b4f9a083f0c8cc015f7b0cc12a3bc88db632610.

This causes some bluestore test failures due to an ENOENT right after a
new object is created.  The simplest reproducer is the
ObjectStore/StoreTest.FiemapEmpty/2 test.

Fixes: https://tracker.ceph.com/issues/42495
Signed-off-by: Sage Weil <sage@redhat.com>

os/bluestore,comon,erasure-code: chmod -x source files

Fixes

*** WARNING: ./usr/src/debug/ceph-15.0.0-6512.g62bd825.el8.x86_64/src/common/obj_bencher.cc is executable but has empty or no shebang, removing executable bit
*** WARNING: ./usr/src/debug/ceph-15.0.0-6512.g62bd825.el8.x86_64/src/erasure-code/shec/determinant.c is executable but has empty or no shebang, removing executable bit
*** WARNING: ./usr/src/debug/ceph-15.0.0-6512.g62bd825.el8.x86_64/src/os/bluestore/AvlAllocator.cc is executable but has empty or no shebang, removing executable bit
*** WARNING: ./usr/src/debug/ceph-15.0.0-6512.g62bd825.el8.x86_64/src/os/bluestore/AvlAllocator.h is executable but has empty or no shebang, removing executable bit

in rpm build.

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

Merge PR #31170 into master

* refs/pull/31170/head:
src/telemetry: remove, now lives in ceph-telemetry.git

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

Merge PR #31131 into master

* refs/pull/31131/head:
crush: remove invalid upmap items

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

Merge PR #31154 into master

* refs/pull/31154/head:
auth/cephx: always initialize local variables

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

Merge PR #31167 into master

* refs/pull/31167/head:
os/bluestore: do not mark per_pool_omap updated unless we fixed it

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

mgr/dashboard: accept socket error 0

see
https://github.com/cherrypy/cheroot/commit/9d73226ab28a07a38cbb6816edf9ff46be4f9de3

Fixes: https://tracker.ceph.com/issues/38378
Signed-off-by: Kefu Chai <kchai@redhat.com>

mgr/dashboard: accept exceptions from builtin SSL

see also https://github.com/cherrypy/cheroot/pull/4

so we don't panic when client is trying to talk with us with an
unsupported protocol, the exception should be accepted, and the
client can fallback to supported protocol.

Fixes: https://tracker.ceph.com/issues/38378
Signed-off-by: Kefu Chai <kchai@redhat.com>

mgr/dashboard: extract monkey patches for cherrypy out

so they are less distracting to the dashboard developers. and probably
less scaring.

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

mgr/diskprediction_cloud: refactor timeout() decorator

* timeout() is never passed any parameter when being called, so let's
  remove the parameters list of "seconds" and "error_message"
* use `getattr()` instead of `hasattr()` for retrieving the
  member variable of `self`
* pass `self` to wrapper function explicitly.
* return `func()` right away.
* hardwire the error message of `TimeoutError` to "Timer expired",
  because
  - as neither errno.ETIME nor errno.ETIMEOUT is portable
  - the only caller of `TimeoutError` is `timeout()`, so there is
    no need to have the flexibility to pass a different error message
* use `wraps()` as a decorator, simpler this way.

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

mon/OSDMonitor: use initializer_list<> for {si,iec}_options

* use initializer_list<> for {si,iec}_options, no need to uset set<>
* remove the comments, the variable names are self-documented.

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

Merge pull request #31010 from pdvian/wip-fix-target-max

mon/OSDMonitor.cc: Allow pool set target_max_(objects/bytes) with SI/IEC units

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

Merge pull request #30434 from smithfarm/wip-41820

qa: enable dashboard tests to be run with "--suite rados/dashboard"

Reviewed-by: Laura Paduano <lpaduano@suse.com>
Reviewed-by: Ernesto Puerta <epuertat@redhat.com>
Reviewed-by: Kefu Chai <kchai@redhat.com>

Merge pull request #31005 from tchaikov/wip/qa/tasks/ceph/cleanup

qa/tasks/ceph.py: remove unused variables

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

qa/tasks/cbt: run stop-all.sh when finishing up

stop-all.sh will work if the right deps are there (currently we lack 'nc')

also killall -9 java to be sure.

Fixes: https://tracker.ceph.com/issues/42496
Signed-off-by: Sage Weil <sage@redhat.com>

Merge PR #31086 into master

* refs/pull/31086/head:
doc/bootstrap.rst: fix github's url

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

Merge PR #31169 into master

* refs/pull/31169/head:
mgr/ssh: implement (synchronous) describe_service
ceph-daemon: ls: replace 'active' bool with 'state' enum
ceph-daemon: include container_id and version in 'ls' output
ceph-daemon: tolerate systemctl is-* exit codes
mgr/ssh: modernize timeout config option

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

doc/bootstrap.rst: fix github's url

Signed-off-by: Alexandre Bruyelles <jack@jack.fr.eu.org>

Merge pull request #31126 from wjwithagen/wjw-fix-std-bitset

rgw: Select the std::bitset to resolv ambiguity

Reviewed-by: Adam C. Emerson <aemerson@redhat.com>
Reviewed-by: Casey Bodley <cbodley@redhat.com>
Reviewed-by: Kefu Chai <kchai@redhat.com>

src/telemetry: remove, now lives in ceph-telemetry.git

Signed-off-by: Dan Mick <dan.mick@redhat.com>

mgr/ssh: implement (synchronous) describe_service

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

ceph-daemon: ls: replace 'active' bool with 'state' enum

('running', 'inactive', 'error', 'unknown')

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

ceph-daemon: include container_id and version in 'ls' output

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

ceph-daemon: tolerate systemctl is-* exit codes

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

mgr/ssh: modernize timeout config option

(and rename it)

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

Merge PR #31141 into master

* refs/pull/31141/head:
ceph-daemon: fix warning message
ceph-daemon: fix enable with conflicting ceph.target files
test_ceph_daemon.sh: use same fsid as qa/standalone/test_ceph_daemon.sh
test_ceph_daemon.sh: use other latest-master image
ceph-daemon: fix pathify (fix shell --config/--keyring)
ceph-daemon: /var/run/ceph -> /var/run/ceph/$fsid

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

ceph-daemon: fix warning message

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

Merge pull request #31164 from tchaikov/wip/doc/rados/fix-typo

doc/rados/deployment/ceph-deploy-mon: fix typo

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

ceph-daemon: fix enable with conflicting ceph.target files

+ make description better

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

os/bluestore: do not mark per_pool_omap updated unless we fixed it

We only fix the per-pool-omap issues if we do a non-shallow fsck.

Fixes: https://tracker.ceph.com/issues/42490
Signed-off-by: Sage Weil <sage@redhat.com>

doc/rados/deployment/ceph-deploy-mon: fix typo

s/comingling/commingling/

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

Merge pull request #31113 from sungjunyoung/do_cmake-warning

do_cmake: Warn user about slow debug performance only for not set

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

Merge PR #31130 into master

* refs/pull/31130/head:
ceph-daemon: only set up crash dir mount if it exists

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

Merge PR #31138 into master

* refs/pull/31138/head:
mon: fix tell to hybrid octopus/pre-octopus mons

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

restful: Use node_id for _gather_leaf_ids

The _gather_leaf_ids function doesn't need the node structure, it only
needs the id.

Signed-off-by: Boris Ranto <branto@redhat.com>

restful: Query nodes_by_id for items

The node dict that is passed to the _gather_leaf_ids function from the
_gather_osds function does not have 'items' in it. We also can't use
buckets at this point since those only exist for leaf nodes, not all
nodes.

We need to query the nodes_by_id dict to get 'items' for a node inside
the _gather_leaf_ids function instead.

Signed-off-by: Boris Ranto <branto@redhat.com>

python-common: Add mypy testing

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

Merge pull request #31006 from zy751713126/fix_features

librbd: fix rbd_features_to_string output

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

Merge pull request #30955 from runsisi/wip-fix-dup-lock

librbd: force reacquire lock if blacklist is disabled

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

Merge pull request #30954 from mxdInspur/thickimg_create_progress

rbd:  creating thick-provision image progress percent info exceeds 100%

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

Merge pull request #31143 from badone/wip-tracker-38466-enable-librabbitmq-devel

ceph.spec.in: enable amqp_endpoint on RHEL8 by default

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

Merge pull request #31110 from tchaikov/wip-cmake/enable-seastar-with-dpdk

cmake: support `Seastar_DPDK=ON` option

Reviewed-by: Casey Bodley <cbodley@redhat.com>
Reviewed-by: Yingxin Cheng <yingxin.cheng@intel.com>

auth/cephx: always initialize local variables

to silence GCC warnings like:

rc/auth/cephx/CephxProtocol.h:309:5: warning: 'type' may be used uninitialized in this function [-Wmaybe-uninitialized]
     if (i != tickets_map.end())
     ^~

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

ceph.spec.in: Enable amqp_endpoint on RHEL8 by default

RHEL/CentOS 8 now provide librabbitmq-devel so we can enable it as a
build requirement.

Fixes: https://tracker.ceph.com/issues/38466
Signed-off-by: Brad Hubbard <bhubbard@redhat.com>

test_ceph_daemon.sh: use same fsid as qa/standalone/test_ceph_daemon.sh

That way these scripts both clean up for each other (and are obviously
bogus clusters).

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

test_ceph_daemon.sh: use other latest-master image

This one is newer?

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

ceph-daemon: fix pathify (fix shell --config/--keyring)

The relative path works on one of my machines but not my laptop; the
full CWD works everywhere.  This is presumably better anyway!

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

ceph-daemon: /var/run/ceph -> /var/run/ceph/$fsid

This is better than having a single /var/run/ceph on the host with a
weird naming scheme.  Among other things, it means that we can access
the asok for any daemon for a given fsid from any container on the same
host with the same fsid (notably, a shell).

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

mon: fix tell to hybrid octopus/pre-octopus mons

We can't decide whether to use the new tell command style based on the
monmap.min_mon_release alone because some mons may be octopus even though
that hasn't updated yet.  The same goes for if we look at the combined
features for the cluster--the underlying problem is the monmap doesn't
tell us which mons are octopus and which ones aren't, so we don't know
how to behave.

Instead, allow octopus+ mons to advertise the converted tell commands
going forward, for compatibility with pre-octopus clients (who do the old
style of tell) and for octopus+ clients talking to a min_mon_release <
octopus cluster.

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

ceph-daemon: only set up crash dir mount if it exists

Sometimes we run containers on a host that doesn't have a crash dir set
up (becuase no daemon has been deployed).  Examples include shell and
ceph-volume.

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

Merge pull request #31133 from dzafman/wip-42476

ceph-objectstore-tool: call collection_bits() crashes on the meta col…

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

Merge pull request #31135 from athanatos/sjust/wip-alien-echo-fix

test_alien_echo: update to use crimson:: namespace

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

Merge PR #31121 into master

* refs/pull/31121/head:
mon/MonClient: fix mon tell to older mons

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

Merge PR #31095 into master

* refs/pull/31095/head:
qa: do not check pg count for new data_isolated volume

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

test/crimson/test_alien_echo: update to use crimson:: namespace

See: bd023211d999f1e2b70f48254507524d0763bd33
Signed-off-by: Samuel Just <sjust@redhat.com>

Merge pull request #30964 from markhpc/wip-bs-cache-trim-pinned

os/bluestore: Keep separate onode cache pinned list.

ceph-objectstore-tool: call collection_bits() crashes on the meta collection

Skip new check for meta collection
test:
    Turn off osd_pool_default_pg_autoscale_mode just like bash tests do
    Fix test by checking for new error message

Caused by: f88b3534549bb2ede76e9f3e8b6b641b5886a9f4

Fixes: https://tracker.ceph.com/issues/42476
Signed-off-by: David Zafman <dzafman@redhat.com>

Merge pull request #31132 from smithfarm/wip-42474

build-integration-branch: take PRs in chronological order

do_cmake: Disable shell command printing feature before cmake execution

Signed-off-by: Junyoung, Sung <junyoung.sung@navercorp.com>

build-integration-branch: change shebang to python3

The word "python" in the shebang means Python 2, but Python 2 is deprecated and
should no longer be used.

Signed-off-by: Nathan Cutler <ncutler@suse.com>

build-integration-branch: take PRs in chronological order

Fixes: https://tracker.ceph.com/issues/42474
Signed-off-by: Nathan Cutler <ncutler@suse.com>

crush: remove invalid upmap items

We can not cancel in verify_upmap if remap an osd to different root bucket,

cluster topology:
osd.0 ~ osd.29 belongs to datacenter 1
osd30 ~ osd.59 belongs to datacenter 2
crush rule:
  take datacenter 1
  chooseleaf 2 host
  emit
  take datacenter 2
  chooseleaf 2 host
  emit
The pg's primary osd in datacenter 1.

We should cancel the pgid from upmap_items like below
1) from [26,12,54,46] to [30,12,54,46]
pg_upmap_items [26,30]

2) from [16,25,53,31] to [40,25,53,0]
pg_upmap_items [16,20,31,0]

Signed-off-by: huangjun <huangjun@xsky.com>

do_cmake: Warn user about slow debug performance only for not set

Signed-off-by: Junyoung, Sung <junyoung.sung@navercorp.com>

rgw: Select the std::bitset to resolv ambiguity

Signed-off-by: Willem Jan Withagen <wjw@digiware.nl>

mon/MonClient: fix mon tell to older mons

For older mons, we need to call into send_command so that we trigger the
compat code that futzes with our mon sessions.

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

Merge PR #30848 into master

* refs/pull/30848/head:
mon/PGMap: drop indentation on 'df' human output

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

cmake: support `Seastar_DPDK=ON` option

seastar does not build any 3rd party libraries. but we do. currently, we
build DPDK/SPDK when they are enabled for the messenger backend and
bluestore backend.

in this change, the same function is reused for preparing the DPDK
libraries if `Seastar_DPDK` is enabled. probably we should outsource
this work to install-deps.sh in future, but before that, let's keep
this as a builtin "feature".

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

cmake: enable Finddpdk to find DPDK to be built

find_package(dpdk...) is used by seastar to find DPDK if Seastar_DPDK is
enabled. but `build_dpdk()` only exposes `dpdk::dpdk`, and it does not
set the variables like `dpdk_INCLUDE_DIRS` and `dpdk_LIBRARIES` when it
gets called. so we need to adapt Finddpdk to query them from `dpdk::dpdk`
target.

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

Merge pull request #31112 from smithfarm/wip-distros-sle

qa/distros: add SLE-12-SP3 and SLE-15-SP1

Reviewed-by: Kyr Shatskyy <kyrylo.shatskyy@suse.de>

qa/distros: add SLE-12-SP3 and SLE-15-SP1

Ceph luminous is known to run on SLE-12-SP3 and nautilus on SLE-15-SP1, so add
these two to qa/distros/all.

Signed-off-by: Nathan Cutler <ncutler@suse.com>

cmake: define dpdk_LIBRARIES properly

dpdk_LIBRARIES should not be a list of dpdk library target, it should be
a list of paths to them. this also align with the definition of
`Finddpdk.cmake` in seastar project. so, if `Seastar_DPDK` is defined,
we should be able to offer the expected `dpdk_LIBRARIES` for seastar, as
our `Finddpdk.cmake` has higher priority than the one in seastar, and is
used when `find_package(dpdk...)` is called.

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

cmake: refactor BuildDPDK

* extract `do_export_dpdk()` into a separated function
* no need to check for `dpdk-ext` target before calling `build_dpdk()`,
  as the name of this target should be hidden from the caller of this
  function.

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

Merge pull request #31069 from tchaikov/wip/crimson/replace-ceph-with-crimson

crimson: s/ceph/crimson/ in namespace names

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

test/crimson: silence deprecated warnings

* do not discard returned future<>
* do not use variadic future<>

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

crimson: s/ceph/crimson/ in namespace names

to help differentiate the symbols shared by classic and crimson osd,
after this change, all crimson code will live in "crimson::" namespaces.
and in a follow-up change, all classic code used by crimson will live in
"ceph::" namespaces.

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

crimson/osd/osd_operation.h: s/ceph_assert/assert/

so we don't need to include `include/ceph_assert.h` for it. and
presumably, `ceph_assert()` is for fail the application early when it is
not able to recover from a failure, while `assert()` is for debugging.

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