> On 10/30/2014 11:40 AM, Cheng Wei-Chung wrote:
>> Dear all:
>>
>> I meet a strange situation. First, I show my ceph status as following:
>>
>> cluster fb155b6a-5470-4796-97a4-185859ca6953
>> ......
>> osdmap e25234: 20 osds: 20 up, 20 in
>> pgmap v2186527: 1056 pgs, 4 pools, 5193 GB data, 1316 kobjects
>> 8202 GB used, 66170 GB / 74373 GB avail
>> 1056 active+clean
>>
>> The replica size of my pool configuration is 2 and 8202 GB is total
>> actual data usage and 5193 GB is about my actual data size.
>> Is that right?
>>
>> I though I have 5193 GB data, at least I should use the 5193 *
>> 2(replica size) = 10386 GB ?
>> Did anyone meet the same situation as me?
>> or I just get something misunderstanding about the data usage in ceph status?
>>
>
> Do all the OSDs have a dedicated filesystem or is there something else
> on those filesystems?
>
yes, I make xfs per OSD and no other data store on those filesystems.
> The OSDs report back to the monitors on how much they used based on what
> "df" tells them.
>
This means I can manually calculate the df return value for the data
usage result about ceph status?
> The data used comes from the Placement Group statistics however.
>
So could I get some methods to check with PG for confirming the
Integrity of my data?
> Can you check with "ceph df" to see in which pools the data is?
>
there are my ceph df return value:
GLOBAL:
SIZE AVAIL RAW USED %RAW USED
74373G 66170G 8202G 11.03
POOLS:
NAME ID USED %USED MAX AVAIL OBJECTS
volumes 5 4734G 6.37 31926G 1289307
images 6 458G 0.62 31926G 58897
data 7 0 0 31926G 0
metadata 8 0 0 31926G 0
If you need more information please let me know.
Thanks!!
>> many thanks!
>> _______________________________________________
>> ceph-users mailing list
>> [email protected]
>> http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
>>
>
>
> --
> Wido den Hollander
> 42on B.V.
> Ceph trainer and consultant
>
> Phone: +31 (0)20 700 9902
> Skype: contact42on
>
>
> _______________________________________________
> ceph-users mailing list
> [email protected]
> http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
>
_______________________________________________
ceph-users mailing list
[email protected]
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
