Hi Tommi ,
Actually I was thinking of "Availability" of related files to
particular host. I wanted to guide ceph in some way to store related
files for host on his local osd so that I/O over network can be
reduced . One more doubt I would like to ask is Whether read/write
requests are sent to primary only Or only write requests are sent to
primary while read request can be served by any OSD having object
replica ?
Regards,
Hemant Surale.
On Tue, Aug 28, 2012 at 8:54 PM, Tommi Virtanen <[email protected]> wrote:
> On Mon, Aug 27, 2012 at 11:50 PM, hemant surale <[email protected]>
> wrote:
>> Hi All ,
>> I was unable to figure out how we can direct the ceph to store
>> data based on host needs i.e. Host A wants to keep frequently used
>> data on its own local OSD. So my question is about how we can direct
>> such write/read requests to local OSD than on network? Is it possible
>> ?
>
> Due to the nature of CRUSH -- being a deterministic pseudorandom
> distribution and all -- the writer does not get to choose where n
> object is stored.
>
> The closest thing you can do is have a RADOS pool with a CRUSH ruleset
> that puts the primary on that machine. As you'd need a pool per
> machine, this really doesn't scale up well.
>
> Perhaps the benefits of the scheme you are envisioning aren't as great
> as you think they would be. Often accessing lots of fast servers over
> a fast network performs better than a single local disk.
--
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
