Project

General

Profile

Bug #39392

[mgr][balancer] Error EAGAIN: Too many objects (0.000000 > 0.000000) are misplaced; try again later

Added by Aleksandr Rudenko 2 months ago.

Status:
New
Priority:
Normal
Assignee:
-
Category:
balancer module
Target version:
Start date:
04/19/2019
Due date:
% Done:

0%

Source:
Community (user)
Tags:
Backport:
Regression:
No
Severity:
3 - minor
Reviewed:
Affected Versions:
ceph-qa-suite:
Pull request ID:

Description

Luminous 12.2.12

ceph -s

  cluster:
    id:     b5c9ceb9-3c3c-440a-8d8a-a1492896d9cc
    health: HEALTH_WARN
            noscrub,nodeep-scrub flag(s) set
            1 pools have many more objects per pg than average

  services:
    mon: 3 daemons, quorum a,b,c
    mgr: b(active), standbys: a, c
    osd: 20 osds: 20 up, 20 in
         flags noscrub,nodeep-scrub
    rgw: 2 daemons active

  data:
    pools:   7 pools, 1216 pgs
    objects: 2.73M objects, 4.46TiB
    usage:   9.24TiB used, 12.6TiB / 21.8TiB avail
    pgs:     1216 active+clean
ceph df

GLOBAL:
    SIZE        AVAIL       RAW USED     %RAW USED
    21.8TiB     12.6TiB      9.24TiB         42.35
POOLS:
    NAME                          ID     USED        %USED     MAX AVAIL     OBJECTS
    .rgw.root                     1      1.09KiB         0       3.59TiB           4
    default.rgw.control           2           0B         0       3.59TiB           8
    default.rgw.meta              3      6.21KiB         0       3.59TiB          30
    default.rgw.log               4           0B         0       3.59TiB         183
    default.rgw.buckets.index     5           0B         0       3.59TiB           5
    default.rgw.buckets.data      6      4.02TiB     52.84       3.59TiB     2606034
    rbd                           7       455GiB     15.69       2.39TiB      125366
ceph osd df

ID CLASS WEIGHT  REWEIGHT SIZE    USE     AVAIL   %USE  VAR  PGS
12   hdd 0.90909  1.00000  931GiB  571GiB  360GiB 61.33 1.45 158
13   hdd 0.90909  1.00000  931GiB  439GiB  492GiB 47.11 1.11 145
14   hdd 0.90909  1.00000  931GiB  438GiB  493GiB 47.02 1.11 142
15   hdd 0.90909  1.00000  931GiB  569GiB  362GiB 61.07 1.44 156
16   hdd 1.81879  1.00000 1.82TiB  599GiB 1.23TiB 32.18 0.76 239
17   hdd 1.81879  1.00000 1.82TiB  339GiB 1.49TiB 18.19 0.43 229
18   hdd 1.81879  1.00000 1.82TiB  471GiB 1.36TiB 25.28 0.60 237
19   hdd 1.81879  1.00000 1.82TiB  592GiB 1.24TiB 31.80 0.75 224
 0   hdd 0.90909  1.00000  931GiB  443GiB  488GiB 47.56 1.12 158
 1   hdd 0.90909  1.00000  931GiB  438GiB  492GiB 47.10 1.11 150
 2   hdd 0.90909  1.00000  931GiB  444GiB  487GiB 47.69 1.13 161
 3   hdd 0.90909  1.00000  931GiB  578GiB  353GiB 62.13 1.47 176
 4   hdd 0.90909  1.00000  931GiB  315GiB  616GiB 33.81 0.80 162
 5   hdd 0.90909  1.00000  931GiB  567GiB  364GiB 60.95 1.44 152
 6   hdd 0.90909  1.00000  931GiB  445GiB  486GiB 47.78 1.13 167
 7   hdd 0.90909  1.00000  931GiB  575GiB  356GiB 61.73 1.46 167
 8   hdd 0.90909  1.00000  931GiB  441GiB  490GiB 47.35 1.12 152
 9   hdd 0.90909  1.00000  931GiB  574GiB  357GiB 61.64 1.46 169
10   hdd 0.90909  1.00000  931GiB  312GiB  619GiB 33.51 0.79 153
11   hdd 0.90909  1.00000  931GiB  313GiB  618GiB 33.64 0.79 159
                    TOTAL 21.8TiB 9.24TiB 12.6TiB 42.35
MIN/MAX VAR: 0.43/1.47  STDDEV: 13.58

ceph balancer status

{
    "active": true,
    "plans": [],
    "mode": "upmap" 
}

Balancer does not work and i have:

ceph balancer optimize default.rgw.buckets.data

Error EAGAIN: Too many objects (0.000000 > 0.000000) are misplaced; try again later

Balancer worked perfectly in past when all OSDs were 1TB. Some times ago i add new 2TB OSDs and balancer stopped working.

Also available in: Atom PDF