Replace `boost::variant` with `std::variant` as part of our effort to reduce
third-party dependencies in favor of C++ standard library alternatives.
Benefits include:
- Improved code readability and maintainability
- Reduced external dependency surface
- More consistent API usage with other components
Ville Ojamo [Sat, 26 Apr 2025 03:54:12 +0000 (10:54 +0700)]
doc/radosgw: Improve formatting in STS.rst
Some text uses hardcoded ordered list numbers and as a result
have duplicate list numbers. Move to use RST ordered lists and
indent the contents of each list item correctly. Fix referenced
list item number in text that pointed to a wrong list item.
Format references to parameter names, user IDs, file names etc
as inline code.
Add prompts to example commands.
Change instances of "rgw" or "Radosgw" in text to "RGW".
Signed-off-by: Ville Ojamo <14869000+bluikko@users.noreply.github.com>
Ville Ojamo [Sat, 26 Apr 2025 04:17:16 +0000 (11:17 +0700)]
doc/radosgw: Fix RST syntax rendeded as text in oidc.rst
Empty line after starting a pre-formatted block with the double-colon
syntax is required, otherwise the double-colon does nothing and is just
rendered as-is as "::" and there would be no following pre-formatted
block.
Add empty lines after the double-colon syntax so that the following
block is rendered pre-formatted.
Also add bash privileged prompts to a block with 2 example CLI commands.
Signed-off-by: Ville Ojamo <14869000+bluikko@users.noreply.github.com>
Ville Ojamo [Sat, 26 Apr 2025 03:30:57 +0000 (10:30 +0700)]
doc/radosgw: Fix length of section header underlines in oidc.rst
Set the length of the RST syntax underlining for section headers to be
as long as the section header text.
Also it seems common to indent the document title with one space at both
the beginning and the end so do that.
Signed-off-by: Ville Ojamo <14869000+bluikko@users.noreply.github.com>
Patrick Donnelly [Fri, 25 Apr 2025 19:00:39 +0000 (15:00 -0400)]
Merge PR #62833 into main
* refs/pull/62833/head:
qa: test charmap changes with dir and snaps
mds: check for snapshots on parent snaprealms
mds: use strict_strtobool for parsing bools
common: take string_view for strict_tobool
this change is created in the same spirit of bb1fa818.
when building the tree with clang-21, following warning was raised:
```
/home/kefu/dev/ceph/src/libcephfs_proxy/proxy_async.c:43:9: warning: arithmetic on a pointer to void is a GNU extension [-Wgnu-pointer-arith]
43 | data += iov->iov_len;
| ~~~~ ^
1 warning generated.
```
this change should address this warning by casting a `void*` pointer to
`char*` pointer before performing arithmetic operation on it.
Ville Ojamo [Fri, 25 Apr 2025 09:00:20 +0000 (16:00 +0700)]
doc/radosgw: Fix section header level in config-ref.rst
The section named "QoS Settings" looks like it should not be a child of
the section "SSE-S3 Settings". Move it to be a sibling to it, on the
same level instead.
Signed-off-by: Ville Ojamo <14869000+bluikko@users.noreply.github.com>
Ville Ojamo [Fri, 25 Apr 2025 07:16:52 +0000 (14:16 +0700)]
doc/radosgw: Fix section header levels in multisite-sync-policy.rst
The section header levels are reversed so the hierarchy in the TOC is
incorrect. Switch around the section header levels to make the TOC
hierarchy correct, for example individual examples are children of the
"Examples" section.
Signed-off-by: Ville Ojamo <14869000+bluikko@users.noreply.github.com>
src/mon/PGMap.cc: check unfound obejcts in `get_unavailable_pg_in_pool_map`
If a pool has any PG with unfound objects, we should consider
it unavailable for the availability score. If a PG has unfound
objects, it will be recorded in PGMap.
In `get_unavailable_pg_in_map`, if a PG has unfound obejcts,
we add it to `pool_pg_unavailable_map`.
When we update the `pg_stat` we don't
check whether the pg state is in `stale`.
Therefore, the attribute `last_unstale`
will always get updated even if the pg
state actually contains `stale`.
Solution:
Place a condition to only update
the attribute `last_unstale` when
we the pg truly doesn't have `stale`
in its state.
Kamoltat [Thu, 26 Oct 2023 19:08:37 +0000 (19:08 +0000)]
src/mon/PGMap.cc: init pool_availability
Added PoolAvailability Struct
Modified PGMap.cc to include a k,v map:
`pool_availability`.
The key being the `poolid` and value
is `PoolAvailability`
Init the function:
`PGMap::get_unavailable_pg_in_pool_map()`
to identify and aggregate all the PGs we
mark as `unavailable` as well as the pool
that associates with the unavailable PG.
Also, included `pool_availability`
to `PGMapDigest::dump()`.
Adam Kupczyk [Mon, 10 Jun 2024 16:03:24 +0000 (16:03 +0000)]
os/bluestore: Add admin socket commands to inspect onode metadata
Add admin socket commands:
1) bluestore collections
Lists collections.
2) bluestore list <coll> [start object] [max count]
Lists collection coll starting from object (optional). Default 100 entries. 0 = unlimited.
3) bluestore onode metadata <object>
Prints onode metadata as seen by BlueStore.
It might happen (usually in tests) that 2 BlueStore instances are created at the same time.
Since admin commands are unique, it fails to register.
Use first register to detect whether we can register at all.
Adam Kupczyk [Tue, 8 Apr 2025 08:36:21 +0000 (08:36 +0000)]
os/bluestore: Add do_write_v2_compressed()
Modify do_write_v2() to branch into do_write_v2_compressed().
Segmented and regular cases are recognized and handled properly.
New do_write_v2_compressed() oversees compression / recompression.
Make one Estimator per Collection.
It makes possible for estimator to learn in collection specific compressibility.
In write_v2_compressed use compressor already selected in choose_write_options.
Make Collection create Estimator on first use.
Adam Kupczyk [Tue, 8 Apr 2025 11:03:22 +0000 (11:03 +0000)]
os/bluestore/compression: Main part of recompression feature
Add feature of recompression scanner that looks around write region to see how much
would be gained, if we read some more around and wrote more.
Added Compression.h / Compression.cc.
Added debug_bluestore_compression dout.
Created Scanner class.
Provides write_lookaround() for scanning loaded extents.
Adam Kupczyk [Wed, 9 Apr 2025 16:03:52 +0000 (16:03 +0000)]
os/bluestore/compression: Estimator class
Add CMake rules to compile.
Add bluestore_compression dout subsys.
Created Estimator class.
It is used by Scanner to decide if specific extent is to be recompressed.
Prepare for future machine learning / adaptive algorithm for estimation.
So far logic of Estimator is relatively simple.
It learns expected recompression values and uses them in next iterations to predict.
Max Kellermann [Thu, 24 Apr 2025 05:17:48 +0000 (07:17 +0200)]
mds/Locker: use ceph_abort_msg() instead of ceph_assert()
This ceph_assert() always fails, but depending on the configuration
value `ceph_assert_supresssions`, execution may continue, but the
`dir` variable is left uninitialized. This leads to a compiler
warning:
/home/jenkins-build/build/workspace/ceph-api/src/mds/Locker.cc:451:22: error: variable 'dir' is used uninitialized whenever 'if' condition is false [-Werror,-Wsometimes-uninitialized]
clang then suggests to nullptr-initialize the variable:
/home/jenkins-build/build/workspace/ceph-api/src/mds/Locker.cc:447:11: note: initialize the variable 'dir' to silence this warning
447 | CDir *dir;
| ^
| = nullptr
This, however, is a very bad idea because all this does is suppress
the warning; it still crashes the process.
Since there's no recovery from this problem, let's switch to
ceph_abort_msg() which is [[noreturn]] and the compiler can deduce
that `dir` is always initialized when it's used.
Signed-off-by: Max Kellermann <max.kellermann@ionos.com>
Merge pull request #62693 from ronen-fr/wip-rf-iocnt
osd/scrub: performance counters for I/O performed by the scrubber
Reviewed-by: Alex Ainscow <aainscow@uk.ibm.com> Reviewed-by: Bill Scales <bill_scales@uk.ibm.com> Reviewed-by: Samuel Just <sjust@redhat.com> Reviewed-by: Radoslaw Zarzynski <rzarzyns@redhat.com>
N Balachandran [Mon, 21 Apr 2025 11:34:08 +0000 (17:04 +0530)]
rbd: display correct mirror state when creating
The mirror image state is set to MIRROR_IMAGE_STATE_CREATING
when the image is first created on the secondary, but was displayed
as "unknown" by the rbd info command. This has been fixed.
Fixes: https://tracker.ceph.com/issues/70963 Signed-off-by: N Balachandran <nithya.balachandran@ibm.com>
HealthMonitor: Add topology-aware netsplit detection and warning
Problem:
Currently, Ceph cannot detect and report network partitions (netsplits)
between monitors in different topology locations in a consolidated way.
While stretch mode can handle partitions through monitor elections,
users lack visibility into the topology-level view of network
disconnections, making troubleshooting difficult.
Solution:
This implementation adds a hierarchical netsplit detection mechanism that:
- Uses DirectedGraph structure for netsplit detection
- Maps monitor disconnections to relevant CRUSH topology levels
- Aggregates individual disconnections into location-level reports when appropriate
- Detects complete location-level netsplits when ALL monitors between locations
cannot communicate
- Reports specific topology locations experiencing complete communication failures
- Falls back to individual monitor-level reporting for partial disconnections
- Handles monitors with missing location data gracefully
- Leverages HealthMonitor::check_for_mon_down to receive a set of down monitors,
efficiently avoiding false netsplit reports for monitors already known to be down
- Implements smart filtering that correctly excludes down monitors from location-based
analysis, ensuring accurate netsplit reporting at both individual and topology levels
The implementation produces user-friendly health warnings:
1. For complete location netsplits: "Netsplit detected between dc1 and dc2"
2. For individual monitor disconnections: "Netsplit detected between mon.a and mon.d"
Performance considerations:
- Time complexity: O(m²) where m is the number of monitors
- Space complexity: O(m²) for connection tracking
- Practical impact is minimal as monitor count is typically small (3-7)
librbd: disallow "rbd trash mv" if image is in a group
Removing an image that is a member of a group has always been
disallowed. However, moving an image that is a member of a group to
trash is currently allowed and this is deceptive -- the only reason for
a user to move an image to trash should be the intent to remove it.
More importantly, group APIs operate in terms of image names -- there
are no corresponding variants that would operate in terms of image IDs.
For example, even though internally GroupImageSpec struct stores an
image ID, the public rbd_group_image_info_t struct insists on an image
name. When rbd_group_image_list() encounters a trashed member image
(i.e. one that doesn't have a name), it just fails with ENOENT and no
listing gets produced at all until the offending image is restored from
trash. Something like this can be very hard to debug for an average
user, so let's make rbd_trash_move() fail with EMLINK the same way as
rbd_remove() does in this scenario.
The one case where moving a member image to trash makes sense is live
migration where the source image gets trashed to be almost immediately
replaced by the destination image as part of preparing migration.
EMLINK is returned by rbd_remove() if the image is a member of a group.
Add a dedicated exception similar to ImageBusy or ImageHasSnapshots and
a test for it.
projects / ceph.git / log