cephadm: convert cephadm agent to a daemon form

The cephadm agent is a bit special in that it will not be converted
to a ContainerDaemonForm (it is not containerized) but we still want
to have it registered as a DeamonForm so that the deamon_type can be
passed to create and have it resolve correctly.

Signed-off-by: John Mulligan <jmulligan@redhat.com>

cephadm: convert SNMPGateway to a ContainerDaemonForm

Signed-off-by: John Mulligan <jmulligan@redhat.com>

cephadm: convert CustomContainer to a ContainerDaemonForm

Signed-off-by: John Mulligan <jmulligan@redhat.com>

cephadm: convert NFSGanesha to a ContainerDaemonForm

Signed-off-by: John Mulligan <jmulligan@redhat.com>

cephadm: add func to deploy any generic ContainerDaemonForm

While there are no ContainerDaemonForms implemented yet, add a function
that uses the ContainerDaemonForm methods to construct a deployment
for the container based daemons.

Signed-off-by: John Mulligan <jmulligan@redhat.com>

cephadm: add ContainerDaemonForm

Add a supplemental DaemonForm subclass that helps deploy container
based daemons in a standard fashion. Most of these methods are
optional and should have sensible defaults.

Signed-off-by: John Mulligan <jmulligan@redhat.com>

cephadm: move DeploymentType to deploy.py

The DeploymentType is used by a number of other classes and functions
and has no dependencies beyond enum and is safe to move.

Signed-off-by: John Mulligan <jmulligan@redhat.com>

cephadm: move firewalld related items to firewalld.py

Signed-off-by: John Mulligan <jmulligan@redhat.com>

cephadm: remove direct daemon-class deps from firewall

Signed-off-by: John Mulligan <jmulligan@redhat.com>

cephadm: move sysctl specific functions to sysctl.py

Signed-off-by: John Mulligan <jmulligan@redhat.com>

cephadm: remove direct daemon-type deps from sysctl

Using the appropriate daemon form we can break the direct dependency
that the sysctl setup function has on particular classes and use
a generic interface.

Signed-off-by: John Mulligan <jmulligan@redhat.com>

cephadm: add test_daemon_form.py

Signed-off-by: John Mulligan <jmulligan@redhat.com>

cephadm: introduce daemon forms to cephadm.py

Introduce the DeamonForm base class to cephadm.py and make various
daemon-type classes into fully fleged deamon form classes.

Some classes already had a semi-standard `init` classmethod for
instantiation. In these cases the new `create` classmethod is a thin
wrapper over the existing method. In cases where the class was not
already being instantiated a minimal set of methods are added.

Signed-off-by: John Mulligan <jmulligan@redhat.com>

cephadm: add daemon_form.py: bases and funcs for daemon forms

Create daemon_form.py containing the DaemonForm class and a few
subclasses and utility functions for working with DaemonForms.
In a future commit, DaemonForm will become the base class for
the current assortment of classes named after the daemon or
family of daemon they help manage.

A daemon form, think "form" as in "template" or "mold", assists
in setting up, creating, and managing daemons controlled with
cephadm. Because cephadm supports a variety of services the
DaemonForm is an abstract base class and the module also supports
additional ABCs that may be used by DaemonForms to implement
optional features.

The daemon forms that are expected to be used directly must be
registered using the provided decorator. This is an explicit extra
step so that common bases that inherit from DaemonForm can be
implemented. Plus explicit is better than implicit. :-)
All DeamonForm subclasses are expected to provide a small set
of standard methods so that the types can be chosen, instantiated,
and used a common manner.

Signed-off-by: John Mulligan <jmulligan@redhat.com>

cephadm: add some unit test coverage for deploying nfs, snmp

Signed-off-by: John Mulligan <jmulligan@redhat.com>

Merge pull request #53436 from rhcs-dashboard/fix-bootstrap-script

mgr/dashboard: fix bootstrap script for cephadm installation

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

Merge pull request #45069 from ifed01/wip-ifed-alloc-stuff

os/bluestore: fixing some non-critical bugs for HybridAllocator

Reviewed-by: Adam Kupczyk <akupzyk@ibm.com>

Merge pull request #53642 from rhcs-dashboard/fix-62964-main

mgr/dashboard: Fix user/bucket count in rgw overview dashboard

Reviewed-by: Pedro Gonzalez Gomez <pegonzal@redhat.com>
Reviewed-by: Ankush Behl <cloudbehl@gmail.com>

Merge pull request #53792 from zdover23/wip-doc-2023-10-03-rados-operations-control-2-of-x

doc/rados: edit ops/control.rst (2 of x)

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

mgr/dashboard: fix bootstrap script for cephadm installation

Fixes: https://tracker.ceph.com/issues/62827
Signed-off-by: avanthakkar <avanjohn@gmail.com>

Merge branch 'main' into wip-doc-2023-10-03-rados-operations-control-2-of-x

Signed-off-by: zdover23 <zac.dover@proton.me>

Merge pull request #53791 from zdover23/wip-doc-2023-10-03-rados-operations-control-1-of-x

doc/rados: edit ops/control.rst (1 of x)

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

doc/rados: edit ops/control.rst (2 of x)

Edit doc/rados/operations/control.rst (2 of x).

Co-authored-by: Cole Mitchell <cole.mitchell.ceph@gmail.com>
Co-authored-by: Anthony D'Atri <anthony.datri@gmail.com>
Signed-off-by: Zac Dover <zac.dover@proton.me>

doc/rados: edit ops/control.rst (1 of x)

Edit doc/rados/operations/control.rst (1 of x).

Co-authored-by: Cole Mitchell <cole.mitchell.ceph@gmail.com>
Co-authored-by: Anthony D'Atri <anthony.datri@gmail.com>
Signed-off-by: Zac Dover <zac.dover@proton.me>

Merge PR #53494 into main

* refs/pull/53494/head:
mds: fix stray CInodes' use-after-free bug when submit ELid entry

Reviewed-by: Venky Shankar <vshankar@redhat.com>
Reviewed-by: Dhairya Parmar <dparmar@redhat.com>
Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>

Merge pull request #53332 from zhsgao/mds_overload_epochs

mds: add option mds_bal_overload_epochs

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

Merge pull request #53746 from rhcs-dashboard/enhance-usage-bar

mgr/dashboard: rbd image hide usage bar when disk usage is not provided

Reviewed-by: Ankush Behl <cloudbehl@gmail.com>
Reviewed-by: Nizamudeen A <nia@redhat.com>

Merge pull request #53307 from rhcs-dashboard/fix-62721-main

mgr/dashboard: Rgw Multi-site naming improvements

Reviewed-by: Pedro Gonzalez Gomez <pegonzal@redhat.com>
Reviewed-by: Nizamudeen A <nia@redhat.com>

Merge pull request #52938 from batrick/mon-loop-refactor

mon: refactor loop variable names

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

Merge pull request #52733 from batrick/i59813

osd/OSDMonitor: check svc is writeable before changing pending

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

Merge pull request #53449 from ljflores/wip-tracker-62833

osd: fix read balancer logic to avoid redundant primary assignment

Merge pull request #53671 from rhcs-dashboard/fix-protect-if-layering-enabled

mgr/dashboard: enable protect option if layering enabled

Reviewed-by: Ankush Behl <cloudbehl@gmail.com>
Reviewed-by: Sunil Angadi <Sunil.Angadi@ibm.com>

Merge pull request #53673 from rhcs-dashboard/fix-62984-main

mgr/dashboard: show a message to restart the rgw daemons after moving from single-site to multi-site

Reviewed-by: Ankush Behl <cloudbehl@gmail.com>
Reviewed-by: Nizamudeen A <nia@redhat.com>

Merge pull request #53703 from adk3798/cephadm-refactor-ssh

cephadm: start ssh.py in cephadmlib

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

Merge PR #53600 into main

* refs/pull/53600/head:
mds/FSMap: allow upgrades if no up mds

Reviewed-by: Milind Changire <mchangir@redhat.com>

mgr/dashboard: enable protect option if layering enabled

Fixes: https://tracker.ceph.com/issues/63076
Signed-off-by: avanthakkar <avanjohn@gmail.com>

mgr/dashboard: show a message to restart the rgw daemons after moving from single-site to multi-site

Fixes: https://tracker.ceph.com/issues/62984
Signed-off-by: Aashish Sharma <aasharma@redhat.com>

Merge pull request #53522 from rhcs-dashboard/landing-page-layout-fix

mgr/dashboard: fix the landing page layout issues

Reviewed-by: Pedro Gonzalez Gomez <pegonzal@redhat.com>
Reviewed-by: Ankush Behl <cloudbehl@gmail.com>
Reviewed-by: Ernesto Puerta <epuertat@redhat.com>

mgr/dashboard: rbd image hide usage bar when disk usage is not provided

Fixes: https://tracker.ceph.com/issues/63037
Signed-off-by: Pedro Gonzalez Gomez <pegonzal@redhat.com>

Merge pull request #53699 from rhcs-dashboard/tls_v1.2_toglle

mgr/dashboard: allow tls 1.2 with a config option

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

Merge pull request #53697 from rhcs-dashboard/mds-number-validator-fix

mgr/dashboard: fix cephfs form validator

Reviewed-by: Pedro Gonzalez Gomez <pegonzal@redhat.com>
Reviewed-by: Aashish Sharma <aasharma@redhat.com>
Reviewed-by: Ankush Behl <cloudbehl@gmail.com>

mgr/dashboard: fix the landing page layout issues

We were following a row-col grid layout for the landing page.
First row includes Details, Status and Capacity
Second row for Inventory and Cluster Utilization

So if one of the item in the first row increases, it pushes the entire
second row downwards.

To fix this, I made a col-row grid.

First col has Details and Inventory in two rows.
Second col has Status and Capacity as a col and Cluster Utilization as a
single row

Fixes: https://tracker.ceph.com/issues/62961
Signed-off-by: Nizamudeen A <nia@redhat.com>
Co-authored-by: cloudbehl <cloudbehl@gmail.com>

Merge PR #53754 into main

* refs/pull/53754/head:
osd: correct unsigned/signed compiler wrn

Reviewed-by: Laura Flores <lflores@redhat.com>
Reviewed-by: Igor Fedotov <ifedotov@suse.com>

osd: correct unsigned/signed compiler wrn

    /home/pdonnell/ceph/src/osd/OSD.cc: In member function ‘void OSD::ShardedOpWQ::stop_for_fast_shutdown()’:
    /home/pdonnell/ceph/src/osd/OSD.cc:11143:41: warning: comparison of integer expressions of different signedness: ‘int’ and ‘uint32_t’ {aka ‘unsigned int’} [-Wsign-compare]
    11143 |   for (int shard_index = 0; shard_index < osd->num_shards; shard_index++) {

Fixes: https://tracker.ceph.com/issues/62851
Fixes: 210dbd4ff19ea66fd2f0109cc15aad53349be52f
Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>

Merge pull request #47148 from NitzanMordhai/wip-nitzan-test-cls-lock-expiered-locks-start

test/cls_lock: expired lock before unlock and start check

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

Merge pull request #53616 from ronen-fr/wip-rf-cnceq

rgw/test: fix compiler warning

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

mds/FSMap: allow upgrades if no up mds

This is to support the fail_fs scenario for cephadm where max_mds >= 1
and all MDS are down.

Fixes: https://tracker.ceph.com/issues/62682
Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>

cephadm: format black cephadmlib/ssh.py

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

cephadm: start ssh.py in cephadmlib

As part of the cephadm refactoring process
to split cephadm into multiple python files,
start "ssh.py" that includes some functions used
for setting up and testing ssh connections,
primarily as part of bootstrap.

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

Merge pull request #53735 from idryomov/wip-63010

qa/suites/krbd: stress test for recovering from watch errors

Reviewed-by: Mykola Golub <mgolub@suse.com>

Merge pull request #50350 from BBoozmen/oozmen_rgw_beast_log_bucket_name

rgw/logging: add bucket and object names to the log line for ops-log

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

Merge pull request #52144 from linuxbox2/wip-nochown-bucket

rgw/file: make setattr(...) a no-op on buckets

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

Merge pull request #52960 from cbodley/wip-62411

rgw/lc: remove_bucket_config() doesn't update xattrs on bucket delete

Reviewed-by Matt Benjamin <mbenjamin@redhat.com>

Merge pull request #53266 from VVoidV/lbr-chunk-upload-buffer-fix

rgw: improve buffer list utilization in the chunkupload scenario

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

Merge pull request #53525 from RickyMaRui/fix-SignatureDoesNotMatch

rgw: fix SignatureDoesNotMatch when extra headers start with 'x-amz'

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

Merge pull request #53602 from cbodley/wip-62938

rgw: s3website doesn't prefetch for web_dir() check

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

Merge pull request #53614 from xrmeng8756/master

rgw: fix rgw rate limiting RGWRateLimitInfo class decode_json max_rea…

Reviewed-by: Jiffin Tony Thottan <jthottan@redhat.com>

Merge pull request #53676 from smanjara/data-sync-fairness-crash

rgw/multisite: call drain before flushing markers in incremental sync

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

Merge pull request #53680 from cbodley/wip-62989

rgw/keystone: EC2Engine uses reject() for ERR_SIGNATURE_NO_MATCH

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

Merge pull request #53684 from cbodley/wip-rgw-keystone-yield

rgw/async: use optional_yield for keystone and kms requests

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

Merge pull request #53691 from max-licc/radosgw_start_watcher_crash

rgw: fix unwatch crash at radosgw startup

Reviewed-by: Adam Emerson <aemerson@redhat.com>

Merge pull request #53714 from cbodley/wip-63004

[CVE-2023-43040] rgw: Fix bucket validation against POST policies

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

Merge pull request #53724 from adk3798/rm-get_unit_name_by_instance

cephadm: remove get_unit_name_by_instance func

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

Merge pull request #53715 from adk3798/other-haproxy-container-version

cephadm: fix haproxy version with certain containers

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

Merge pull request #53744 from zdover23/wip-doc-2023-10-02-architecture-16-of-x

doc/architecture: edit "Calculating PG IDs"

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

qa/suites/krbd: stress test for recovering from watch errors

Fixes: https://tracker.ceph.com/issues/63010
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

Merge pull request #53606 from Patty8122/control-reaches-end-fix

test: corrected control reaches end by adding a return

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

doc/architecture: edit "Calculating PG IDs"

Edit the section "Calcluating PG IDs" in doc/architecture.rst.

Signed-off-by: Zac Dover <zac.dover@proton.me>

test: corrected control reaches end by adding a return

Signed-off-by: Patty8122 <divyapattisapu@uchicago.edu>

Merge pull request #53580 from AliMasarweh/wip-alimasa-persistent-q-migration

RGW | Bucket Notification: migrating old entries to support persistency control

Reviewed-by: yuvalif <ylifshit@redhat.com>

Merge pull request #46692 from pdvian/wip-unreachable-osd

osd: Report health error if OSD public address is not within subnet

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

Merge pull request #52509 from ericqzhao/main

mgr: add throttle policy for DaemonServer

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

Merge pull request #53740 from zdover23/wip-doc-2023-09-30-architecture-15-of-x

doc/architecture: edit several sections

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

Merge pull request #53618 from adk3798/cephadm-refactor-more-net-utils

cephadm: move more funcs into net_utils.py

Reviewed-by: John Mulligan <jmulligan@redhat.com>
Reviewed-by: Michael Fritch <mfritch@suse.com>

doc/architecture: edit several sections

Edit the following sections in doc/architecture.rst:

 1. Dynamic Cluster Management
 2. About Pools
 3. Mapping PGs to OSDs

The tone of "Dynamic Cluster Management" remains a bit too close to the
tone of marketing material, in my opinion, but I will return to firm it
up when I have finished a once-over of architecture.rst.

Co-authored-by: Anthony D'Atri <anthony.datri@gmail.com>
Signed-off-by: Zac Dover <zac.dover@proton.me>

Merge pull request #53702 from adk3798/cephadm-refactor-registry-login

cephadm: move registry_login to container_engines.py

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

Merge pull request #53617 from adk3798/cephadm-refactor-more-data-utils

cephadm: move more funcs into data_utils.py

Reviewed-by: John Mulligan <jmulligan@redhat.com>
Reviewed-by: Michael Fritch <mfritch@suse.com>

Merge pull request #53612 from adk3798/fix-refreshed-column

mgr/cephadm: fix REFRESHED column of orch ps being unpopulated

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

cephadm: re-format black data_utils.py

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

cephadm: move more funcs into data_utils.py

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

Merge pull request #53682 from phlogistonjohn/jjm-cephadm-move-pathify

cephadm: move pathify & get_file_timestamp to file_utils

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

Merge pull request #53683 from phlogistonjohn/jjm-cephadm-move-getters

cephadm: move context getters

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

Merge pull request #53726 from zdover23/wip-doc-2023-09-29-architecture-14-of-x

doc/architecture: edit "Replication"

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

Merge pull request #53605 from sayantani11/osd-1

script/backport-resolve-issue: Update script with latest versions

rgw: fix rgw rate limiting RGWRateLimitInfo class decode_json max_read_bytes and max_write_bytes field mismatch

Fixes: https://tracker.ceph.com/issues/62955
Signed-off-by: xiangrui meng <mengxr@chinatelecom.cn>

Merge pull request #53723 from zdover23/wip-doc-2023-09-29-architecture-13-of-x

doc/architecture: edit "Data Scrubbing"

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

doc/architecture: edit "Replication"

Edit "Replication" in the "Smart Daemons Enable Hyperscale" section of
doc/architecture.rst.

Signed-off-by: Zac Dover <zac.dover@proton.me>

doc/architecture: edit "Data Scrubbing"

Edit the "Data Scrubbing" listitem in the list of benefits conferred by
the use by OSDs of the aggregate power of the cluster, in the section
"Smart Daemons Enable Hyperscale" in doc/architecture.rst.

Co-authored-by: Anthony D'Atri <anthony.datri@gmail.com>
Signed-off-by: Zac Dover <zac.dover@proton.me>

Merge PR #53594 into main

* refs/pull/53594/head:
qa: fix "no orch backend set" in nfs suite

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

Merge pull request #53707 from zdover23/wip-doc-2023-09-28-architecture-12-of-x

doc/architecture: edit "OSD Membership and Status"

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

cephadm: remove get_unit_name_by_instance func

As it is one line, quite simple, and only
had a single caller, it was decided we'd remove
this function as part of the cephadm refactor.

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

doc/architecture: edit "OSD Membership and Status"

Edit "OSD Membership and Status" in the "Smart Daemons Enable
Hyperscale" section of doc/architecture.rst.

Co-authored-by: Anthony D'Atri <anthony.datri@gmail.com>
Signed-off-by: Zac Dover <zac.dover@proton.me>

cephadm: fix haproxy version with certain containers

Some builds of haproxy containers' output
from "haproxy -v" start with

HAProxy version

rather than

HA-Proxy version

no reason on our end not to accept both

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

rgw: Fix bucket validation against POST policies

It's possible that user could provide a form part as a part of a POST
object upload that uses 'bucket' as a key; in this case, it was
overriding what was being set in the validation env (which is the real
bucket being modified). The result of this is that a user could actually
upload to any bucket accessible by the specified access key by matching
the bucket in the POST policy in said POST form part.

Fix this simply by setting the bucket to the correct value after the
POST form parts are processed, ignoring the form part above if
specified.

Fixes: https://tracker.ceph.com/issues/63004
Signed-off-by: Joshua Baergen <jbaergen@digitalocean.com>

Merge pull request #53608 from cbodley/wip-62536

qa/rgw/sts: keycloak task installs java manually

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

rgw/file: make setattr(...) a no-op on buckets

Shallow fix for apparent unstable behavior after nfs "chown" on
an RGW bucket via RGW NFS.  While we allow buckets to be created
(and subject to ordinary rules, deleted), chown against a bucket
hasn't been tested and potentially is not valid.  Prevent it
altogether for now--if permissions would allow it, chown will
succeed but won't have any effect.

Fixes: https://tracker.ceph.com/issues/61689
Signed-off-by: Matt Benjamin <mbenjamin@redhat.com>

RGW | Bucket Notification: migrating old entries to support persistency control

Signed-off-by: Ali Masarwa <ali.saed.masarwa@gmail.com>

qa/suites/krbd: rename singleton to singleton-msgr-failures

A "singleton without msgr-failures" is wanted in the next commit.

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

Merge pull request #53432 from liu-chunmei/seastore_device_size

crimson/vstart: default seastore_device_size will be out of space f…

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

cephadm: re-format black container_engines.py

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