mgr/cephadm: Add disk rescan feature to the orchestrator
Reviewed-by: Adam King <adking@redhat.com>
Reviewed-by: Anthony D'Atri <anthonyeleven@users.noreply.github.com>
mgr/cephadm: support for miscellaneous config files for daemons
Reviewed-by: Anthony D'Atri <anthonyeleven@users.noreply.github.com>
Reviewed-by: John Mulligan <jmulligan@redhat.com>
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>
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>
With ./box.py --engine docker you can specify you want to use docker
instead of podman. With docker box.py command should be run with sudo.
Signed-off-by: Pere Diaz Bou <pdiazbou@redhat.com>
This addresses a typo that Anthony D'Atri pointed out in
PR#47032. It didn't get addressed there, so it's being
addressed here.
Signed-off-by: Zac Dover <zac.dover@gmail.com>
This PR adds a section to the Developer Guide chapter
"Essentials" that explains what Dependabot is. This
section is adapted from an email from Ernesto Puerta
to the CLT that was sent on 08 Jul 2022.
Co-authored-by: Ernesto Puerta <epuertat@redhat.com>>
Signed-off-by: Zac Dover <zac.dover@gmail.com>
This PR improves the English in the "CephFS Delayed
Deletion" chapter of the Ceph Internals Guide (which
is stored in the doc/dev/ directory).
There was no pressing need to rewrite this. I just saw
that it could be tightened up, and I had the fifteen
minutes I needed to do it.
Co-author: Anthony D'Atri <anthony.datri@gmail.com>
Signed-off-by: Zac Dover <zac.dover@gmail.com>
doc/dev/crimson: Correct the link to Scylla
Reviewed-by: Radoslaw Zarzynski <rzarzyns@redhat.com>
Reviewed-by: Anthony D'Atri <anthony.datri@gmail.com>
Even though box is a rootless podman container, there are some
capabilities that are need like SYS_ADMIN, NET_ADMIN and SYS_TIME.
In rootless podman chronyd is not capabable of using adjtimex syscall
so we had to add -x flag to chronyd entrypoint which states it will
not try to call that function.
To test run `./box.py -v cluster start`.
To fix: There are some problems with /dev permissions so osds cannot be
deployed still. `./box.py cluster start --expanded` won't fully work.
Details:
- Dockerfile:
- Fedora image instead of centos.
- Tons of basic packages aren't installed, therefore, those are
explicitly installed.
- Some container folders are created. It isn't clear/tested whether
those are needed.
- .box_container is an empty file to identify if a command is ran on
containers (see box.py for usage).
- box.py
- podman-compose is missing some features but it compensates with args
as --podman-run-args so you can add some extra commands whenever
podman-compose calls podman run.
- sudo ./box.py cluster start will exit since it can be dangerous.
- Added some sed trickeroo to modify chronyd options. (I don't know
how can I pass OPTIONS to it :P)
- host.py
- setup_ssh now is need on every box container.
- util.py
- rice command printing :3.
- run multiple commands with a multiline string.
Signed-off-by: Pere Diaz Bou <pdiazbou@redhat.com>
Rationale: There are many threshold limits for split and
merge in this doc that just says like:
"A directory fragment is eligible for splitting
when its size exceeds `mds_bal_split_size`
(default 10000)". Need to clarify what 10000 actually
means. This applies to all other such entries in this
doc.
Signed-off-by: Dhairya Parmar <dparmar@redhat.com>
Create the initial mClock QoS params at CONF_DEFAULT level using
set_val_default(). This allows switching to a custom profile on a
running OSD and to make necessary changes to the desired QoS params.
Note that Switching to ‘custom’ profile and then subsequently changing
the QoS params using “config set osd.n …” will be at a higher level i.e.
at CONF_MON.
But When switching back to a built-in profile, the new values won’t take
effect since CONF_DEFAULT < CONF_MON. For the values to take effect, the
config keys created as part of the ‘custom’ profile must be removed from
the ConfigMonitor store after switching back to a built-in profile.
- Added a couple of standalone tests to exercise the scenario.
- Updated the mClock configuration document and the mClock internal
documentation with a couple of typos relating to the best effort weights.
- Added new sections to the mClock configuration document outlining the
steps to switch between the built-in and custom profile and vice-versa.
Fixes: https://tracker.ceph.com/issues/55153
Signed-off-by: Sridhar Seshasayee <sseshasa@redhat.com>
