[Gluster-devel] AFR between two bricks over 3000 miles
krishna at zresearch.com
Thu Mar 27 10:46:41 UTC 2008
We have committed code in afr for write() call, i.e afr returns as soon
as it finds out that one of the writes have succeeded. But we can not
do this for all calls, imagine for open() call there are chances that
we can get write() call and open() has not yet been done on one of
the subvols. We are planning for a new translator called "optimist"
translator which handles these kind of complications when trying
to do the calls in background.
You could see if the change to write() call has improved performance
in your case.
We have not set a timeline for the new translator yet.
On Wed, Mar 26, 2008 at 3:29 AM, <nathan at robotics.net> wrote:
> On Tue, 11 Mar 2008, Krishna Srinivas wrote:
> > Hi Nathan,
> > We are working on a solution, it will take time. Will let you know
> > after it is done, if it is not done with in 2 weeks please follow it
> > up with us.
> Just wanted to touch base and see if there is any progress on this?
More information about the Gluster-devel