|
|
ceph status % S/ E+ y) t6 i( q3 |+ N5 k& M
cluster:% G4 X" s( _" w1 u! S
id: ed60a380-c9bd-11eb-ad78-000c29ad02de
( @, w- m* v% t: |7 o7 D, t9 F2 C health: HEALTH_WARN6 |3 K/ S; g6 j. u* t5 @2 V4 ]% G$ l& n
Degraded data redundancy: 1 pg undersized
( M; {5 [, ?9 {) ]% E" N 6 T6 |) Y( `3 l T3 y, T$ \( ~
services:
( r$ ]" v2 H' F0 z! X7 Q mon: 1 daemons, quorum controller (age 18m)9 }9 J& {# u% _, ?; ~
mgr: controller.ifotuc(active, since 7m)+ W- a, t w- k. d/ F2 G4 b& e
osd: 6 osds: 6 up (since 12m), 6 in (since 12m); 1 remapped pgs! Y" j4 [; `% [% i1 f+ s
4 C% u: |" @' x/ I data:+ H7 w& g/ `2 D0 {
pools: 1 pools, 1 pgs& C8 o9 ^5 }$ ?6 m6 n
objects: 0 objects, 0 B. G5 _% k) N7 X( a0 Z
usage: 6.0 GiB used, 114 GiB / 120 GiB avail
6 E6 U Q- `5 k/ F$ s" q pgs: 1 active+undersized+remapped
8 r& @+ l" s1 b& |0 TDegraded data redundancy: 1 pg undersized9 \7 g1 G- f6 R! O/ S. x
! k+ ]' p% F' B" U$ ^
( k& q/ Q7 }/ k& s5 Z6 L# N: y3 g6 ^6 Y _
将调小size和min_size. j; M4 h( i$ b* w0 t: ?7 A; _
ceph osd pool ls detail / N# S0 u+ L, V, ]3 C' C8 i$ a
pool 1 'device_health_metrics' replicated size 3 min_size 1 crush_rule 0 object_hash rjenkins pg_num 1 pgp_num 1 autoscale_mode on last_change 51 flags hashpspool stripe_width 0 pg_num_min 1 application mgr_devicehealth6 @8 O+ g$ L+ S7 s9 Z+ w& z
3 |) ~; a+ ?5 z/ g, ~ceph osd pool set device_health_metrics size 24 r, a& p9 V& J! h6 T
: a( i* \% |5 _, l. q7 g. Y3 A
ceph osd pool set device_health_metrics min_size 1$ ?$ {4 }- Y2 F e; h* }1 {8 z
set pool 1 min_size to 16 M# N% N$ y/ ?) y+ u# ~' u$ `& p* g
# y2 s. {$ G- ~# K. Q4 q
[root@controller ~]# ceph -s* r" T7 {: J" |
cluster:
/ g1 ]* q U/ S: P& f id: ed60a380-c9bd-11eb-ad78-000c29ad02de# [/ t; k5 t9 o0 n/ \* \
health: HEALTH_OK* a- s5 S- X% w/ M: g
* N9 r7 a7 N7 N$ I: X1 c services:
+ `2 R* R# m: c" e mon: 1 daemons, quorum controller (age 17m)9 r5 U6 ^* [ b1 Y
mgr: controller.ifotuc(active, since 5m)6 C3 [) C+ ^; V8 e
osd: 6 osds: 6 up (since 11m), 6 in (since 11m); 1 remapped pgs: V7 C$ m {" ~( R( f( \& P$ Q
7 P5 K6 [- x: D0 t data:3 v- ?$ H0 p9 [
pools: 1 pools, 1 pgs
: M" f/ r/ ?- m7 C s3 b" M2 J objects: 0 objects, 0 B
/ I" R* M, n! B( I# N: B usage: 6.0 GiB used, 114 GiB / 120 GiB avail5 z1 c3 q, s7 p& L* [" _+ }! {
pgs: 1 active+undersized+remapped; A6 f- H( J9 z: c
! {# S* W) y/ A `1 Q: `8 A: @$ q4 F# N& \7 `
[root@controller ~]# ceph -s
& g8 P, D* D0 K4 F2 t" k1 v cluster:+ f% K2 V% O( s( Z6 v4 r$ n
id: ed60a380-c9bd-11eb-ad78-000c29ad02de/ o# x: K+ }% [& b6 {/ l* u
health: HEALTH_OK P$ l8 T5 w' B( Z3 \/ ?$ z9 z
8 R7 X" W% ]9 V. g
services:9 g4 Y& k. [# P, \, J5 N0 L- v2 o
mon: 1 daemons, quorum controller (age 21m)
% t* o9 ^- h& v+ {) Y mgr: controller.ifotuc(active, since 9m)
# L, T3 ]3 |; Z0 n( B- {' m osd: 6 osds: 6 up (since 15m), 6 in (since 15m); 1 remapped pgs
5 a5 D2 g' z! ?/ D+ W
: P, u1 D. L) @6 F& m/ h data:( g! o) ~. h/ n4 s
pools: 1 pools, 1 pgs
* ~8 g9 r$ D. j0 b, q3 E" J6 H objects: 0 objects, 0 B6 d) {( Q+ n9 ]( b
usage: 6.0 GiB used, 114 GiB / 120 GiB avail
# c/ o) @9 @! w% m& r6 G7 |+ _ pgs: 1 active+clean+remapped$ p* N W& v( i5 {5 e" ^: ~9 U
- e/ ?% l% _" s5 B
设置低于默认值就正常了。奇怪的问题。这是octopus版本的问题。% Y, [0 G Z' s+ R
device_health_metrics 是由cephadm安装后默认生成的。
; B& J l4 R' R% j$ Y+ ?; i( i |
|
发表于 2021-6-10 17:01:51