The cephfs_metadata pool makes sense on ssd, but it won't need a lot of space. Chances are that you'll have plenty of ssd storage to spare for other uses.
Personally, I'm migrating away from a cache tier and rebuilding my OSDs. I am finding that performance with Bluestore OSDs with the block.db on SSDs performs better in most cases than a cache tier, and it is a simpler design. There's some good notes here on good and bad use cases for cache tiering: http://docs.ceph.com/docs/master/rados/operations/cache-tiering/
On Mon, Jun 3, 2019 at 3:35 PM Daniele Riccucci <devster@xxxxxxxxxx> wrote:
Hello,
sorry to jump in.
I'm looking to expand with SSDs on an HDD cluster.
I'm thinking about moving cephfs_metadata to the SSDs (maybe with device
class?) or to use them as cache layer in front of the cluster.
Any tips on how to do it with ceph-ansible?
I can share the config I currently have if necessary.
Thank you.
Daniele
On 17/04/19 17:01, Sinan Polat wrote:
> I have deployed, expanded and upgraded multiple Ceph clusters using ceph-ansible. Works great.
>
> What information are you looking for?
>
> --
> Sinan
>
>> Op 17 apr. 2019 om 16:24 heeft Francois Lafont <francois.lafont.1978@xxxxxxxxx> het volgende geschreven:
>>
>> Hi,
>>
>> +1 for ceph-ansible too. ;)
>>
>> --
>> François (flaf)
>> _______________________________________________
>> 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
>
_______________________________________________
ceph-users mailing list
ceph-users@xxxxxxxxxxxxxx
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
--
Shawn Iverson, CETL
![]()
![]()
![Cybersecurity]()
Director of Technology
Rush County Schools
765-932-3901 option 7
_______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
