Weired. Maybe you can check the source code (src/mon/PGMonitor.cc,
around L1434).
But, looks like there is another command "ceph pg dump_json {all |
summary | sum | pools | ...} which you can try.
On Fri, May 16, 2014 at 2:56 PM, Cao, Buddy <buddy.cao at intel.com> wrote:
> In my env, "ceph pg dump all -f json" only returns result below,
>
> {"version":45685,"stamp":"2014-05-15 23:50:27.773608","last_osdmap_epoch":13875,"last_pg_scan":13840,"full_ratio":"0.950000","near_full_ratio":"0.850000","pg_stats_sum":{"stat_sum":{"num_bytes":151487109145,"num_objects":36186,"num_object_clones":0,"num_object_copies":72372,"num_objects_missing_on_primary":0,"num_objects_degraded":5716,"num_objects_unfound":0,"num_read":8502912,"num_read_kb":611729970,"num_write":2737247,"num_write_kb":340122861,"num_scrub_errors":39,"num_shallow_scrub_errors":39,"num_deep_scrub_errors":0,"num_objects_recovered":267486,"num_bytes_recovered":1120311874505,"num_keys_recovered":94},"stat_cat_sum":{},"log_size":952236,"ondisk_log_size":952236},"osd_stats_sum":{"kb":19626562368,"kb_used":296596996,"kb_avail":19329965372,"hb_in":[],"hb_out":[],"snap_trim_queue_len":0,"num_snap_trimming":0,"op_queue_age_hist":{"histogram":[1,2,0,1,1,2,56,5,20,18],"upper_bound":1024},"fs_perf_stat":{"commit_latency_ms":55456,"apply_latency_ms":408}},"pg_stats_delta":{"stat_sum":{"num_bytes":45867008,"num_objects":10,"num_object_clones":0,"num_object_copies":20,"num_objects_missing_on_primary":0,"num_objects_degraded":0,"num_objects_unfound":0,"num_read":385,"num_read_kb":29360,"num_write":296,"num_write_kb":61320,"num_scrub_errors":0,"num_shallow_scrub_errors":0,"num_deep_scrub_errors":0,"num_objects_recovered":0,"num_bytes_recovered":0,"num_keys_recovered":0},"stat_cat_sum":{},"log_size":0,"ondisk_log_size":0}}
>
> But "ceph pg dump summary" returns much more info than this and even includes a lot pg summaries likes below. Any idea? thanks
> 3.3 30 0 0 0 125829120 77 77 active+clean 2014-05-15 23:41:47.950780 13875'77 13875:129 [11,4] [11,4] 0'0 2014-05-15 20:02:58.937561 0'0 2014-05-13 20:02:53.011326
> 4.4 49 0 0 0 205520896 3001 3001 active+clean 2014-05-15 19:52:41.534026 13875'5784 13875:29766 [41,26] [41,26] 13244'5438 2014-05-15 09:47:09.501469 0'0 2014-05-14 09:45:29.872229
> 5.5 0 0 0 0 0 0 0 active+clean 2014-05-15 19:52:40.344205 0'0 13875:1702 [25,33] [25,33] 0'0 2014-05-15 09:48:33.067498 0'0 2014-05-14 09:47:40.406650
> 6.6 0 0 0 0 0 0 0 active+clean 2014-05-15 09:49:45.360999 0'0 13875:1678 [24,9] [24,9] 0'0 2014-05-15 09:49:45.360951 0'0 2014-05-14 09:47:42.871263
> 7.7 0 0 0 0 0 0 0 active+clean 2014-05-15 20:12:03.812881 0'0 13875:58 [33,25] [33,25] 0'0 2014-05-15 20:12:02.859031 0'0 2014-05-15 20:12:02.859031
> pool 0 0 0 0 0 0 0 0
> pool 1 21 0 0 0 9470 21 21
> pool 2 0 0 0 0 0 0 0
> pool 3 13006 0 2032 0 54429261968 39905 39905
> pool 4 22725 0 3616 0 95251489157 909458 909458
> pool 5 3 0 0 0 262 4 4
> pool 6 0 0 0 0 0 0 0
> pool 7 0 0 0 0 0 0 0
> pool 8 0 0 0 0 0 0 0
> sum 35755 0 5648 0 149680760857 949388 949388
> osdstat kbused kbavail kb hb in hb out
> 0 79744 486985484 487065228 [1,8,9,16,17,32,33,34,35,36,37,47] []
> 1 80068 486985160 487065228 [0,2,8,9,16,17,32,33,34,35,36,37] []
> 2 9602980 477462248 487065228 [1,3,12,13,18,19,38,39,40,41] []
> 3 18896156 468169072 487065228 [2,4,11,12,13,18,19,20,38,39,40,41] []
> 4 15758800 471306428 487065228 [3,5,12,13,18,19,20,38,39,40,41] []
> 5 13118956 473946272 487065228 [4,6,7,11,12,18,20,38,39,40] []
> 6 83000 486982228 487065228 [5,7,14,15,22,23,42,43,44,45,46,47] []
>
>
> Wei Cao (Buddy)
>
> -----Original Message-----
> From: xan.peng [mailto:xanpeng at gmail.com]
> Sent: Friday, May 16, 2014 2:20 PM
> To: Cao, Buddy
> Cc: ceph-users at ceph.com
> Subject: Re: ?ceph pg dump summary ?f json? question
>
> Looks like "ceph pg dump all -f json" = "ceph pg dump summary".
>
> On Fri, May 16, 2014 at 1:54 PM, Cao, Buddy <buddy.cao at intel.com> wrote:
>> Hi there,
>>
>> ?ceph pg dump summary ?f json? does not returns data as much as ?ceph
>> pg dump summary?, are there any ways to get the fully Json format
>> data for ?ceph pg dump summary??
>>
>>
>>
>>
>>
>> Wei Cao (Buddy)
>>
>>
>>
>>
>> _______________________________________________
>> ceph-users mailing list
>> ceph-users at lists.ceph.com
>> http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
>>
