common/options/osd.yaml.in: Change mclock max sequential bandwidth for SSDs
The osd_mclock_max_sequential_bandwidth_ssd is changed to 1200 MiB/s as
a reasonable middle ground considering the broad range of SSD capabilities.
This allows the mClock's cost model to extract the SSDs capability
depending on the cost of the IO being performed.
osd/: Retain the default osd_max_backfills limit to 1 for mClock
The earlier limit of 3 was still aggressive enough to have an impact on
the client and other competing operations. Retain the current default
for mClock. This can be modified if necessary after setting the
osd_mclock_override_recovery_settings option.
Samuel Just [Tue, 11 Apr 2023 15:10:04 +0000 (08:10 -0700)]
osd/scheduler/mClockScheduler: avoid limits for recovery
Now that recovery operations are split between background_recovery and
background_best_effort, rebalance qos params to avoid penalizing
background_recovery while idle.
Samuel Just [Thu, 6 Apr 2023 05:57:42 +0000 (22:57 -0700)]
osd/: differentiate scheduler class for undersized/degraded vs data movement
Recovery operations on pgs/objects that have fewer than the configured
number of copies should be treated more urgently than operations on
pgs/objects that simply need to be moved to a new location.
Samuel Just [Tue, 4 Apr 2023 23:34:17 +0000 (23:34 +0000)]
osd/scheduler: simplify qos specific params in OpSchedulerItem
is_qos_item() was only used in operator<< for OpSchedulerItem. However,
it's actually useful to see priority for mclock items since it affects
whether it goes into the immediate queues and, for some types, the
class. Unconditionally display both class_id and priority.
osd: Retain overridden mClock recovery settings across osd restarts
Fix an issue where an overridden mClock recovery setting (set prior to
an osd restart) could be lost after an osd restart.
For e.g., consider that prior to an osd restart, the option
'osd_max_backfill' was successfully set to a value different from the
mClock default. If the osd was restarted for some reason, the
boot-up sequence was incorrectly resetting the backfill value to the
mclock default within the async local/remote reservers. This fix
ensures that no change is made if the current overriden value is
different from the mClock default.
Modify an existing standalone test to verify that the local and remote
async reservers are updated to the desired number of backfills under
normal conditions and also across osd restarts.
osd: Set default max active recovery and backfill limits for mClock
Client ops are sensitive to the recovery load and must be carefully
set for osds whose underlying device is HDD. Tests revealed that
recoveries with osd_max_backfills = 10 and osd_recovery_max_active_hdd = 5
were still aggressive and overwhelmed client ops. The built-in defaults
for mClock are now set to:
Previously, setting default configs from the configured profile was
split across:
- enable_mclock_profile_settings
- set_mclock_profile - sets mclock_profile class member
- set_*_allocations - updates client_allocs class member
- set_profile_config - sets profile based on client_allocs class member
This made tracing the effect of changing the profile pretty challenging
due passing state through class member variables.
Instead, define a simple profile_t with three constexpr values
corresponding to the three profiles and handle it all in a single
set_config_defaults_from_profile() method.
osd: Modify mClock scheduler's cost model to represent cost in bytes
The mClock scheduler's cost model for HDDs/SSDs is modified and now
represents the cost of an IO in terms of bytes.
The cost parameters, namely, osd_mclock_cost_per_io_usec_[hdd|ssd]
and osd_mclock_cost_per_byte_usec_[hdd|ssd] which represent the cost
of an IO in secs are inaccurate and therefore removed.
The new model considers the following aspects of an osd to calculate
the cost of an IO:
- osd_mclock_max_capacity_iops_[hdd|ssd] (existing option)
The measured random write IOPS at 4 KiB block size. This is
measured during OSD boot-up using OSD bench tool.
- osd_mclock_max_sequential_bandwidth_[hdd|ssd] (new config option)
The maximum sequential bandwidth of of the underlying device.
For HDDs, 150 MiB/s is considered, and for SSDs 750 MiB/s is
considered in the cost calculation.
The following important changes are made to arrive at the overall
cost of an IO,
1. Represent QoS reservation and limit config parameter as proportion:
The reservation and limit parameters are now set in terms of a
proportion of the OSD's max IOPS capacity. The earlier representation
was in terms of IOPS per OSD shard which required the user to perform
calculations before setting the parameter. Representing the
reservation and limit in terms of proportions is much more intuitive
and simpler for a user.
2. Cost per IO Calculation:
Using the above config options, osd_bandwidth_cost_per_io for the osd is
calculated and set. It is the ratio of the max sequential bandwidth and
the max random write iops of the osd. It is a constant and represents the
base cost of an IO in terms of bytes. This is added to the actual size of
the IO(in bytes) to represent the overall cost of the IO operation.See
mClockScheduler::calc_scaled_cost().
3. Cost calculation in Bytes:
The settings for reservation and limit in terms a fraction of the OSD's
maximum IOPS capacity is converted to Bytes/sec before updating the
mClock server's ClientInfo structure. This is done for each OSD op shard
using osd_bandwidth_capacity_per_shard shown below:
The above result is updated within the mClock server's ClientInfo
structure for different op_scheduler_class operations. See
mClockScheduler::ClientRegistry::update_from_config().
The overall cost of an IO operation (in secs) is finally determined
during the tag calculations performed in the mClock server. See
crimson::dmclock::RequestTag::tag_calc() for more details.
4. Profile Allocations:
Optimize mClock profile allocations due to the change in the cost model
and lower recovery cost.
5. Modify standalone tests to reflect the change in the QoS config
parameter representation of reservation and limit options.
Fixes: https://tracker.ceph.com/issues/58529 Fixes: https://tracker.ceph.com/issues/59080 Signed-off-by: Samuel Just <sjust@redhat.com> Signed-off-by: Sridhar Seshasayee <sseshasa@redhat.com>
osd: update PGRecovery queue item cost to reflect object size
Previously, we used a static value of osd_recovery_cost (20M
by default) for PGRecovery. For pools with relatively small
objects, this causes mclock to backfill very very slowly as
20M massively overestimates the amount of IO each recovery
queue operation requires. Instead, add a cost_per_object
parameter to OSDService::awaiting_throttle and set it to the
average object size in the PG being queued.
Fixes: https://tracker.ceph.com/issues/58606 Signed-off-by: Samuel Just <sjust@redhat.com> Signed-off-by: Sridhar Seshasayee <sseshasa@redhat.com>
osd: update OSDService::queue_recovery_context to specify cost
Previously, we always queued this with cost osd_recovery_cost which
defaults to 20M. With mclock, this caused these items to be delayed
heavily. Instead, base the cost on the operation queued.
Fixes: https://tracker.ceph.com/issues/58606 Signed-off-by: Samuel Just <sjust@redhat.com> Signed-off-by: Sridhar Seshasayee <sseshasa@redhat.com>
osd/osd_types: use appropriate cost value for PullOp
See included comments -- previous values did not account for object
size. This causes problems for mclock which is much more strict
in how it interprets costs.
Fixes: https://tracker.ceph.com/issues/58607 Signed-off-by: Samuel Just <sjust@redhat.com> Signed-off-by: Sridhar Seshasayee <sseshasa@redhat.com>
osd/osd_types: use appropriate cost value for PushReplyOp
See included comments -- previous values did not account for object
size. This causes problems for mclock which is much more strict
in how it interprets costs.
Fixes: https://tracker.ceph.com/issues/58529 Signed-off-by: Samuel Just <sjust@redhat.com> Signed-off-by: Sridhar Seshasayee <sseshasa@redhat.com>
Laura Flores [Mon, 1 May 2023 16:28:54 +0000 (16:28 +0000)]
mgr: add urllib3==1.26.15 to mgr/requirements.txt
We do not depend on any particular version of
urllib3, but as a workaround to the incompatibility
of urllib3 constraints between kubernetes and
requests, we need to pin it temporarily to
the version both are happy with.
Fixes: https://tracker.ceph.com/issues/59591 Signed-off-by: Laura Flores <lflores@redhat.com>
(cherry picked from commit 80d460005e44649191aa862fa78bd278644b5237)
Nizamudeen A [Thu, 27 Apr 2023 11:24:24 +0000 (16:54 +0530)]
mgr/dashboard: fix the rbd mirroring configure check
In one-way mirroring, the condition we are checking now for configuring
the mirroring will fail because only one cluster needs to have the
mirror daemon present. Thus even if mirroring is successfuly happening
the page won't load. For now relaxing the rule until we find a better
api call to check for the status
Edit the "stretch mode" section in doc/rados/operations/stretch-mode.rst
so that the procedure is formatted as a procedure and the sentences
correctly have heads.
Co-authored-by: Anthony D'Atri <anthony.datri@gmail.com> Signed-off-by: Zac Dover <zac.dover@proton.me>
(cherry picked from commit a19ff7a5ea9bbd24365648a90abfa1b720c5b231)
Patrick Donnelly [Tue, 25 Apr 2023 14:54:52 +0000 (10:54 -0400)]
Merge PR #50779 into quincy
* refs/pull/50779/head:
mds: add config to decide whether to mark dentry bad
qa: add missing scan_links step for data scan recovery
qa/tasks/cephfs: test damage to dentry's first is caught
qa/tasks/cephfs: use rank_asok and allow specifying rank
qa/tasks: allow specifying timeout command prefix to ceph
mds: provide test configs for creating first corruption
mds: catch damage to dentry's first field
mds: add debugging for pre_cow_old_inode
mds: cleanup code
mds: check for some dentry damage in scrub
mds: remove unused method
mds: note damaged dentry with first gt last
mds: cluster log scrub failure for dirfrag
mds: mark dirfrag good if repaired
mds: only dump past_parent_snap if non-empty
Patrick Donnelly [Tue, 25 Apr 2023 14:53:55 +0000 (10:53 -0400)]
Merge PR #50777 into quincy
* refs/pull/50777/head:
log: fix stderr handling on Windows
log: add tests for stderr writes to fifos
log: use non-blocking atomic writes to stderr fifos
log: invalidate m_fd on close
log: reorg header
Patrick Donnelly [Tue, 25 Apr 2023 14:50:56 +0000 (10:50 -0400)]
Merge PR #50767 into quincy
* refs/pull/50767/head:
qa: simplify and use correct recovery procedure
doc: update alternate meta pool recovery
tools/cephfs/DataScan: add debugging for directory injection
Mohan Sharma [Tue, 27 Dec 2022 06:01:04 +0000 (11:31 +0530)]
ceph-volume: fix drive-group issue
The drive-group expects the batch_args to be a string,
however in the current version it is passed as a list
of one element, thus calling the first item of the list solves the issue.
`api.get_single_pv(filters={'lv_uuid': lv.lv_uuid})` needs to be called
only if `--destroy` is passed in order to remove vg and pv when there's
nothing left.
With old deployments, it is possible that a lv_uuid matches more than 1
PV.
Given that `get_single_pv()` is only needed when `--destroy` is passed,
let's move this call where it is actually needed.
doc/start: edit first 50 lines of documenting-ceph
Edit the first 150 lines of doc/start/documenting-ceph.rst. This is part
of an initiative to harvest the fruits of Cephalocon 2023, at which
documentation proved to be in demand to a surprising degree.
Co-authored-by: Anthony D'Atri <anthony.datri@gmail.com> Signed-off-by: Zac Dover <zac.dover@proton.me>
(cherry picked from commit dd37f94aa4f1de947b1eaf5d82cc529925f5823e)
Line-edit doc/rados/user-management.rst (2 of x). Some internal
references had to be removed, but these will be repaired when the next
part of this file is updated in a future PR.
Co-authored-by: Anthony D'Atri <anthony.datri@gmail.com> Signed-off-by: Zac Dover <zac.dover@proton.me>
(cherry picked from commit e3575bb72f307a27d49fedf3692ca661e3d613a5)
projects / ceph.git / log