Referenced purge subcommand info via ceph osd command label.
Fixes: https://tracker.ceph.com/issues/36605
Signed-off-by: James McClune <jmcclune@mcclunetechnologies.net>
* refs/pull/24270/head:
osd: make 'cache drop' command require 'executable' permission
osd: rename 'drop cache' and 'get cache stats' to group them by component
doc: add documentation for 'drop cache' and 'get cache stats'
osd: don't print osdmap cache stats in 'get cache stats' command
osd: do not clear osdmap cache on 'drop cache' command
osd: offload dumping cache stats to the object store
osd: pass a stream to flush_cache commands for more verbosity
osd: implement flush_cache() method for Filestore
osd: add clear_cache and get_cache_object_count commands
Reviewed-by: Gregory Farnum <gfarnum@redhat.com>
Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
- some cleanup (e.g., use $ms throughput)
- behave if the local host is in the $hosts list (use $ms.remote)
- be clear about updating all mons
- mon.0 -> mon.foo
Signed-off-by: Sage Weil <sage@redhat.com>
- those old pg settings are no longer lost as they go in the OSDMap now
- add a note about pools that were mid-create
Signed-off-by: Sage Weil <sage@redhat.com>
Updated crush map tunables link in rados/configuration/mon-config-ref.rst
Fixes: https://tracker.ceph.com/issues/36339
Signed-off-by: James McClune <jmcclune@mcclunetechnologies.net>
Updated cluster map reference link in rados/troubleshooting/troubleshooting-mon.rst
Fixes: https://tracker.ceph.com/issues/36336
Signed-off-by: James McClune <jmcclune@mcclunetechnologies.net>
as the latest SPDK deprecated and removed spdk_pci_get_device(),
we cannot use the PCI device's serial number as its identifier and the
filename of the file representing the device. in this change, we
are using the PCI device's selector instead, it is also used by SPDK as
part of the transport id.
Tested-by: Tone Zhang <tone.zhang@arm.com>
Signed-off-by: Kefu Chai <kchai@redhat.com>
* refs/pull/24109/head:
doc: update docs for device management
mgr: make devicehealth always-on
mgr/devicehealth: do not create metrics pool on get-device-metrics
mgr/devicehealth: converge OPTIONS and DEFAULTS
mgr/devicehealth: squelch health warnings for unused devices
mgr/devicehealth: show-health-metrics -> get-health-metrics
Reviewed-by: John Spray <john.spray@redhat.com>
Move this out of mgr section and into rados operations.
Leave a placeholder for failure prediction, which is still a work in
progress.
Signed-off-by: Sage Weil <sage@redhat.com>
The rbd-mirror daemon will require permission to read config-keys
from the "rbd/mirror/" prefix. These new profiles (and the new
associated bootstrap-rbd-mirror user) provide the required
permissions.
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
Fixed reference link for hit set type value. Restructured wording in description.
Fixes: https://tracker.ceph.com/issues/34539
Signed-off-by: James McClune <jmcclune@mcclunetechnologies.net>
Also:
- Do not print **offset** until specified
- Count missing objects correctly (used to be primary's local missing)
Signed-off-by: xie xingguo <xie.xingguo@zte.com.cn>
Note that for the keyring we still accept (but ignore) auid lines so that
we can parse old keyrings that have them.
Signed-off-by: Sage Weil <sage@redhat.com>
Correct a number of spelling mistakes and word omissions in the cluster
configuration section of the docs.
Signed-off-by: Bryan Stillwell <bstillwell@godaddy.com>
this helps user to find the balancer feature, and
also addresses the FTBFS of doc like
Warning, treated as error:
/home/jenkins-build/build/workspace/ceph-pr-docs/doc/rados/operations/balancer.rst:document
isn't included in any toctree
this doc was moved from doc/mgr/balancer.rst to
doc/rados/operations/balancer.rst in
a78d600e6b
Signed-off-by: Kefu Chai <kchai@redhat.com>
Restructured OSD event descriptions for better clarification.
Added a link to the write() system call reference.
Signed-off-by: James McClune <jmcclune@mcclunetechnologies.net>
doc: document sizing for `block.db`
Reviewed-by: Ben England <bengland@redhat.com>
Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Reviewed-by: Sage Weil <sage@redhat.com>
I found a couple misspelled words in the crush-map documentation, and also
tweaked the formatting of the CRUSH hierarchy diagram to center some of the
entries.
Signed-off-by: Bryan Stillwell <bstillwell@godaddy.com>
* refs/pull/22433/head:
common/config: Add description to (near)full ratio settings
Reviewed-by: David Zafman <dzafman@redhat.com>
Reviewed-by: Sage Weil <sage@redhat.com>
Also drop all release-specific upgrading instructions (they only go up to
Firefly, anyway - none of the current releases are covered).
Note that all of this verbiage I am removing here can still be accessed
on docs.ceph.com via e.g. http://docs.ceph.com/docs/firefly/
Signed-off-by: Nathan Cutler <ncutler@suse.com>
Performance evaluations of medium to large size Ceph clusters have
demonstrated negligible performance impact from unnecessarily deep
directory hierarchies but significant performance impact from filestore
split and merge activity. Disable merges by default.
Fixes: http://tracker.ceph.com/issues/24686
Signed-off-by: Douglas Fuller <dfuller@redhat.com>
For many users it is not clear that these only apply to initial creation
of the cluster and that afterwards these commands need to be used:
ceph osd set-nearfull-ratio <float[0.0-1.0]>
ceph osd set-full-ratio <float[0.0-1.0]>
Signed-off-by: Wido den Hollander <wido@42on.com>
Remove confusing references to ancient releases. Be a bit more clear
about what signatures do/don't protect.
Signed-off-by: Sage Weil <sage@redhat.com>
Change pg state names according to osd/osd_types.cc as this is what
ceph -s and the prometheus exporter will present to users.
Signed-off-by: Jan Fajerski <jfajerski@suse.com>
OSD::scrub_load_below_threshold divides the 1 min load by the number of
online cpus. The documentation said it would use the load without
division by number of cpus.
Signed-off-by: Dirk Sarpe <dirk.sarpe@idiv.de>
System attributes shown as "object_info", "snapset" and "hashinfo"
Only output user attributes as "attrs"
Drop leading undescore "_" for user attribute keys
Improve logic as to when to show user attributes or specific system attributes
Signed-off-by: David Zafman <dzafman@redhat.com>
The "Config file section names" section was duplicating some of the content
that is now covered better by the "Configuration sections" section.
Signed-off-by: Nathan Cutler <ncutler@suse.com>
- remove a bunch of outdated info
- update to reflect new monitor-based configuration
- move a bunch of common settings to another document/page
Signed-off-by: Sage Weil <sage@redhat.com>
* refs/pull/20688/head:
doc: Drop the output of pg query
doc: Update Monitoring OSDs and PGs
Reviewed-by: Lenz Grimmer <lgrimmer@suse.com>
Reviewed-by: Sage Weil <sage@redhat.com>
Drop the output of pg query from the doc, as it is:
* Very old. The doc has only one-fourth of the latest 'pg query' output.
* Updating the latest 'pg query' output in the doc would be huge.
* too difficult to maintain it in the doc and make it sync with the changes of the actual 'pg query' output.
However, we can insert parts of the output in the doc if necessary.
Signed-off-by: Jos Collin <jcollin@redhat.com>
The EC 3+2 example used to show 40% overhead where it should be 67%. This might confuse new EC user to calculate the correct EC 3+2 profile overhead.
Signed-off-by: Charles Alva <charlesalva@gmail.com>
Includes new test case
Caused by: 5f58301a13
This changed attr consistency checking to exclude system keys,
which required snapset to be handled just like object info.
Fixes: http://tracker.ceph.com/issues/22996
Signed-off-by: David Zafman <dzafman@redhat.com>
Doc says 1000000, but the real figure is two orders of magnitude less:
Option("log_max_recent", Option::TYPE_INT, Option::LEVEL_ADVANCED)
.set_default(500)
.set_daemon_default(10000)
Signed-off-by: Nathan Cutler <ncutler@suse.com>
The config file paths listed in the docs don't match the config file
paths actually searched by ceph (see `src/common/config.cc`). Update the
docs to reflect the current reality.
Signed-off-by: Adam Wolfe Gordon <awg@digitalocean.com>
hit_set parameters aren't described in a sequence they introduced.
This patch effecrively changes the sequence to improve
description's logic.
Signed-off-by: Alexey Stupnikov <aleksey.stupnikov@gmail.com>
Since kraken, Ceph enforces a 1:1 correspondence between CRUSH ruleset and
CRUSH rule, so effectively ruleset and rule are the same thing, although
the term "ruleset" still survives - notably in the CRUSH rule itself, where it
effectively denotes the number of the rule.
This commit updates the documentation to more faithfully reflect the current
state of the code.
Fixes: http://tracker.ceph.com/issues/20559
Signed-off-by: Nathan Cutler <ncutler@suse.com>
* refs/pull/17678/head:
mon/AuthMonitor: improve error message
mon/OSDMonitor: disallow "all" as a key or value name
cephfs, mon/AuthMonitor, OSD/osdcap: make 'all' a synonym for '*'
vstart.sh: Create an admin user for each CephFS
mon/AuthMonitor: Allow * wildcard for filesystem name
OSD/OSDCap: Allow namespace and pool tag to be combined
OSD/OSDCap: Namespace globbing
mon/AuthMonitor: Use new osd auth caps for ceph fs authorize
OSD/auth caps: Add OSD auth caps based on pool tag
mon/FSCommands: Tag pools used for cephfs by default
mon/OSDMonitor: Add key/value arguments for pool tagging
Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
Define the string 'all' to be a synonym for the wildcard '*'. This
avoids confusion in the event that some auth caps (typically with
ceph fs authorize) are not quoted and thus '*' is expanded by the shell.
Signed-off-by: Douglas Fuller <dfuller@redhat.com>
Cleaner and easier. Also implicitly documents
the config set command, which hadn't been explicitly
called out in the docs before.
Signed-off-by: John Spray <john.spray@redhat.com>
Permit the wildcard * at the end of namespace names to match any
namespace starting with the given prefix. The wildcard is only
allowed at the end of the namespace name.
Example:
allow rw namespace=foo*
Signed-off-by: Douglas Fuller <dfuller@redhat.com>
