Project

General

Profile

Feature #38211

mgr/dashboard: Add custom dialogue for configuring PG scrub parameters

Added by Lenz Grimmer 7 months ago. Updated 2 months ago.

Status:
Resolved
Priority:
Normal
Category:
dashboard/osds
Target version:
Start date:
02/06/2019
Due date:
% Done:

0%

Source:
Tags:
dashboard, management
Backport:
nautilus
Reviewed:
Pull request ID:

Description

Ceph provides multiple configuration options to tweak the behavior of scrub operations, e.g.

  • osd_max_scrubs - the maximum number of simultaneous scrub operations on a given OSD
  • osd_scrub_during_recovery - Allow scrubbing when PGs on the OSD are undergoing recovery
  • osd_scrub_begin_hour, osd_scrub_end_hour - the hours of day (0 to 24) that define a time window when the scrubbing can happen
  • osd_scrub_begin_week_day, osd_scrub_end_week_day - the days of week (0 to 6) that define a time window when the scrubbing can happen
  • osd_scrub_min_interval, osd_scrub_max_interval - Scrub each PG no more/no less often than this interval
  • osd_deep_scrub_interval - Deep scrub each PG (i.e., verify data checksums) at least this often
  • osd_scrub_auto_repair_num_errors - Maximum number of detected errors to automatically repair

While it's possible to modify these via the config editor, it may be worthwhile providing a custom "Scrub configuration" dialogue, similar to the recovery profiles or the global OSD settings.
Are there any other settings that should be added here?


Related issues

Copied to mgr - Backport #40059: nautilus: mgr/dashboard: Add custom dialogue for configuring PG scrub parameters Resolved

History

#1 Updated by Tatjana Dehler 6 months ago

Lenz Grimmer wrote:

Are there any other settings that should be added here?

By 'settings' you mean any other config options, don't you?

I had a look at the config options page and found 32 OSD options where 'scrub' is mentioned:

  • osd_debug_deep_scrub_sleep -> Inject an expensive sleep during deep scrub IO to make it easier to induce preemption
  • osd_deep_scrub_interval -> Deep scrub each PG (i.e., verify data checksums) at least this often
  • osd_deep_scrub_keys -> Number of keys to read from an object at a time during deep scrub
  • osd_deep_scrub_large_omap_object_key_threshold -> Warn when we encounter an object with more omap keys than this
  • osd_deep_scrub_large_omap_object_value_sum_threshold -> Warn when we encounter an object with more omap key bytes than this
  • osd_deep_scrub_randomize_ratio -> Scrubs will randomly become deep scrubs at this rate (0.15 -> 15% of scrubs are deep)
  • osd_deep_scrub_stride -> Number of bytes to read from an object at a time during deep scrub
  • osd_deep_scrub_update_digest_min_age -> Update overall object digest only if object was last modified longer ago than this
  • osd_max_scrubs -> Maximum concurrent scrubs on a single OSD
  • osd_op_queue_mclock_scrub_lim -> mclock weight of limit requests
  • osd_op_queue_mclock_scrub_res -> mclock reservation of scrub requests
  • osd_op_queue_mclock_scrub_wgt -> mclock weight of scrub requests
  • osd_requested_scrub_priority
  • osd_scrub_auto_repair -> Automatically repair damaged objects detected during scrub
  • osd_scrub_auto_repair_num_errors -> Maximum number of detected errors to automatically repair
  • osd_scrub_backoff_ratio -> Backoff ratio for scheduling scrubs
  • osd_scrub_begin_hour -> Restrict scrubbing to this hour of the day or later
  • osd_scrub_begin_week_day -> Restrict scrubbing to this day of the week or later
  • osd_scrub_chunk_max -> Maximum number of objects to scrub in a single chunk
  • osd_scrub_chunk_min -> Minimum number of objects to scrub in a single chunk
  • osd_scrub_cost -> Cost for scrub operations in work queue
  • osd_scrub_during_recovery -> Allow scrubbing when PGs on the OSD are undergoing recovery
  • osd_scrub_end_hour -> Restrict scrubbing to hours of the day earlier than this
  • osd_scrub_end_week_day -> Restrict scrubbing to days of the week earlier than this
  • osd_scrub_interval_randomize_ratio -> Ratio of scrub interval to randomly vary
  • osd_scrub_invalid_stats
  • osd_scrub_load_threshold -> Allow scrubbing when system load divided by number of CPUs is below this value
  • osd_scrub_max_interval -> Scrub each PG no less often than this interval
  • osd_scrub_max_preemptions -> Set the maximum number of times we will preempt a deep scrub due to a client operation before blockinient IO to complete the scrub
  • osd_scrub_min_interval -> Scrub each PG no more often than this interval
  • osd_scrub_priority -> Priority for scrub operations in work queue
  • osd_scrub_sleep -> Duration to inject a delay during scrubbing

#2 Updated by Tatjana Dehler 6 months ago

  • Assignee set to Tatjana Dehler

#3 Updated by Lenz Grimmer 6 months ago

  • Status changed from New to In Progress

#4 Updated by Lenz Grimmer 5 months ago

  • Target version set to v15.0.0
  • Backport set to nautilus

#5 Updated by Tatjana Dehler 5 months ago

  • Pull request ID set to 27072

#6 Updated by Patrick Seidensal 3 months ago

  • Status changed from In Progress to Need Review

#7 Updated by Lenz Grimmer 3 months ago

  • Status changed from Need Review to Pending Backport
  • Affected Versions v14.0.0, v14.2.0, v14.2.1, v15.0.0 added

#8 Updated by Nathan Cutler 3 months ago

  • Copied to Backport #40059: nautilus: mgr/dashboard: Add custom dialogue for configuring PG scrub parameters added

#9 Updated by Lenz Grimmer 2 months ago

  • Status changed from Pending Backport to Resolved

Also available in: Atom PDF