On 06/09/2015 11:10 AM, Jeff Darcy wrote:
So, maybe passing these options as a mount command doesn't work/is a
no-op, but what I don't understand is why -- given that there is no
measure by which glusterfs should ever conclude the replica in the
"other" region is ever faster than the replica in the "same" region.
If read-subvolume or read-subvolume-index is somehow not getting through,
then we're back to the read-hash-mode default - which does *not* try to
use round-trip-time measurements. Still, it should give different
results for different files.
In
fact, it appears as though glusterfs is *preferring* the slower replica.
It's hard to see how that would be the case, since the code to set
read_child based on first-to-reply seems to be *missing* in the current
code. :( What version are you running, again?
Sorry for neglecting to mention the version, it's 3.7.1.
_______________________________________________
Gluster-users mailing list
Gluster-users@gluster.org
http://www.gluster.org/mailman/listinfo/gluster-users
