Hi Ker,
Unfortunately no. Ceph uses the journal for internal consistency and
atomicity and it can't use the XFS journal for it. On the BTRFS side,
we've been investigating allowing the Ceph journal to be on the same
disk as the OSD and doing a clone() operation to effectively reduce the
journal write penalty, but that feature hasn't been implemented yet.
Mark
On 07/11/2013 08:12 PM, ker can wrote:
Hi,
Is it possible to turn off ceph journaling if I switch to xfs ?
For using it as a storage layer for hadoop we're concerned about the
additional requirements for separate SSDs ($$) etc. In our testing
we're seeing a performance hit when using the same disk for both journal
+ data ... so we're investigating the possibility of being able to turn
it off completely, since xfs is already a journaled FS.
thanks
kc
_______________________________________________
ceph-users mailing list
ceph-users@xxxxxxxxxxxxxx
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
_______________________________________________
ceph-users mailing list
ceph-users@xxxxxxxxxxxxxx
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
