On Thursday, April 21, 2016, Benoît LORIOT <benoit.lor...@aevoo.fr> wrote:

> Hello,
>
> we want to disable readproxy cache tier but before doing so we would like
> to make sure we won't loose data.
>
> Is there a way to confirm that flush actually write objects to disk ?
>
> We're using ceph version 0.94.6.
>
>
> I tried that, with cephfs_data_ro_cache being the hot storage pool and
> cephfs_data being the cold storage pool
>
> # rados -p cephfs_data_ro_cache ls
>
> then choose a random object in the list : 100004d6142.00000000
>
> Find the object on cache disk :
>
> # ceph osd map cephfs_data_ro_cache 100004d6142.00000000
> osdmap e301 pool 'cephfs_data_ro_cache' (6) object '100004d6142.00000000'
> -> pg 6.d010000 (6.0) -> up ([4,5,8], p4) acting ([4,5,8], p4)
>
> Object is in the pg 6.0 on OSD 4, 5 and 8, I can find the file on disk.
>
> # ls -l
> /var/lib/ceph/osd/ceph-4/current/6.0_head/DIR_0/DIR_0/DIR_0/DIR_0/100004d6142.00000000__head_0D010000__6
> -rw-r--r--. 1 root root 0 Apr  8 19:36
> /var/lib/ceph/osd/ceph-4/current/6.0_head/DIR_0/DIR_0/DIR_0/DIR_0/100004d6142.00000000__head_0D010000__6
>
> Flush the object :
>
> # rados -p cephfs_data_ro_cache cache-try-flush 100004d6142.00000000
>
> Find the object on disk :
>
> # ceph osd map cephfs_data 100004d6142.00000000
> osdmap e301 pool 'cephfs_data' (1) object '100004d6142.00000000' -> pg
> 1.d010000 (1.0) -> up ([1,7,2], p1) acting ([1,7,2], p1)
>
> Object is in the pg 1.0 on OSD 1, 7 and 2, I can't find the file on disk
> on any of the 3 OSDs
>
> # ls -l
> /var/lib/ceph/osd/ceph-1/current/1.0_head/DIR_0/DIR_0/DIR_0/DIR_0/100004d6142.*
> ls: cannot access
> /var/lib/ceph/osd/ceph-1/current/1.0_head/DIR_0/DIR_0/DIR_0/DIR_0/100004d6142.*:
> No such file or directory
>
>
>
> What am I doing wrong ? To me it seems that nothing is actually flushed to
> disk.
>
>
Is the directory actually hashed four levels deep? Or is it shallower and
you're looking too far down?
-Greg


> Thank you,
> Ben.
>
_______________________________________________
ceph-users mailing list
ceph-users@lists.ceph.com
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com

Reply via email to