Project

General

Profile

Actions
Copy link

Bug #41054

open

disable "pgs not deep-scrubbed in time" message

Added by hoan nv over 4 years ago. Updated 12 months ago.

Status:
New
Priority:
Normal
Assignee:
-
Category:
-
Target version:
-
% Done:

0%

Source:
Tags:
Backport:
Regression:
No
Severity:
3 - minor
Reviewed:
Affected Versions:
v14.2.2
ceph-qa-suite:
Pull request ID:
Crash signature (v1):
Crash signature (v2):

Description

Hi all.

I has 2 ceph cluster run on 14.2.2 versions, all osds have options

osd_deep_scrub_stride = 1048576
osd_deep_scrub_interval = 2592000
osd_scrub_min_interval = 604800
osd_scrub_max_interval = 1209600

ceph health

health: HEALTH_WARN
9757 pgs not deep-scrubbed in time

I added two options and restart all ceph-mon

mon_warn_pg_not_deep_scrubbed_ratio = 0
mon_warn_pg_not_scrubbed_ratio = 0

But message still exist.

I think the options mon_warn_pg_not_deep_scrubbed_ratio = 0 will remove this message but not.

How i can disable this message.

Thanks

Actions
Copy link
 #1

Updated by Sten Aus over 4 years ago

Exatctly same experience here.

Actions
Copy link
 #2

Updated by Sten Aus about 4 years ago

@hoan nv
have you tried newer version of Ceph to see if this still exists?

Actions
Copy link
 #3

Updated by hoan nv about 4 years ago

Sten Aus wrote:

@hoan nv
have you tried newer version of Ceph to see if this still exists?

I have config on ceph mon

mon_warn_pg_not_deep_scrubbed_ratio = 0
mon_warn_pg_not_scrubbed_ratio = 0

then i restart ceph mon this message gone. it works.

Actions
Copy link
 #4

Updated by Sten Aus about 4 years ago

I've added this to config file and tried using injectargs on both commands

ceph tell 'mon.*' injectargs '--mon_warn_pg_not_deep_scrubbed_ratio 0'
ceph tell 'mon.*' injectargs '--mon_warn_pg_not_scrubbed_ratio 0'

and restarted monitor service, but that did not help.
Nor did it help after I've edited this to a greater timeperiod.

Actions
Copy link
 #5

Updated by Sten Aus about 4 years ago

I was able to get a workaround with

ceph osd pool set <poolname> deep_scrub_interval 2592000
ceph osd pool set <poolname> scrub_min_interval 604800
ceph osd pool set <poolname> scrub_max_interval 1209600
Actions
Copy link
 #6

Updated by Katarzyna Myrek almost 4 years ago

Sten Aus, try to set it on the mgrs:

ceph config set mgr mon_warn_pg_not_deep_scrubbed_ratio 0
ceph config set mgr mon_warn_pg_not_scrubbed_ratio 0

in the config file use global or mgr section.

Actions
Copy link
 #7

Updated by can zhu 12 months ago

if update the config, need to restart the mgr instance? and other question is, when running "ceph config", how to know set the params to mon or osd or mgr?

Actions
Copy link
 #8

Updated by hoan nv 12 months ago

can zhu wrote:

if update the config, need to restart the mgr instance? and other question is, when running "ceph config", how to know set the params to mon or osd or mgr?

No, don't need restart mgr.

You can see prefix of param,

Eg : mon_warn_pg_not_deep_scrubbed_ratio is param to mon

if not mon,mgr,osd,rgw it is a global param

Actions
Copy link
 #9

Updated by can zhu 12 months ago

after add "mon_warn_pg_not_deep_scrubbed_ratio" to the ceph.conf, not active; but update "mon_warn_pg_not_deep_scrubbed_ratio" trought the 'ceph config' cmd, it's active.

Actions
Copy link
 #10

Updated by hoan nv 12 months ago

can zhu wrote:

after add "mon_warn_pg_not_deep_scrubbed_ratio" to the ceph.conf, not active; but update "mon_warn_pg_not_deep_scrubbed_ratio" trought the 'ceph config' cmd, it's active.

after add this config to ceph.conf, you must restart ceph mon to apply config.

Actions
Copy link

Also available in: Atom PDF