[Gluster-devel] AFR between two bricks over 3000 miles
lidaobing at gmail.com
Thu Feb 28 15:34:33 UTC 2008
On Thu, Feb 28, 2008 at 11:33 PM, <nathan at robotics.net> wrote:
> On Thu, 28 Feb 2008, LI Daobing wrote:
> > On Thu, Feb 28, 2008 at 2:07 PM, Anand Avati <avati at zresearch.com> wrote:
> >> on the client. you also might want to put this write-behind + io-cache pair
> >> in the subvolume path of afr which leads towards the remote site alone. Also
> >> make sure afr does not have the remote site as the first subvolume, and has
> >> the option read-subvolume <local-volume> so that reads are not scheduled to
> >> the remote site.
> > in 1.3.8pre1, the order of afr's subvolume should be same among
> > clients, otherwise there is a race condition on create and other
> > operations on inode. consider two client are creating a same file at a
> > same time.
> So then it is not possible to have clients each write to their local part
> of the AFR since the order needs to be the same?
afr need write to every sub volume, so you can't only write to the
local part in any case.
More information about the Gluster-devel