Le 08/03/2016 10:11, Krutika Dhananjay a écrit :
IIUC, you would essentially be unmounting the volume from the old building and mounting it near the new building, correct?
Well yes.
In that case, you would need to set the read-subvolume-index command line option appropriately when mounting again.You're right but when mounts are from fstab the volumes may be still busy at end of boot-time. There are solutions: update fstab before shutdown, mount volumes later, re-reboot after editing… But being able to propagate a new local value to client config without umount/mount may be usefull.The selection of the new read child will not happen automatically.
-- Y.
HTH, KrutikaOn Fri, Mar 4, 2016 at 5:59 PM, Yannick Perret <[email protected] <mailto:[email protected]>> wrote:Le 04/03/2016 13:12, Krutika Dhananjay a écrit :You could configure read-subvolume for individual clients using the following mount option: mount -t glusterfs -o "xlator-option=*replicate*.read-subvolume-index=0" <hostname>:/<volname> <mountpoint> In the example above, 0 represents the zeroth child. If a particular client wants to use the second brick as the read child, use the following: mount -t glusterfs -o "xlator-option=*replicate*.read-subvolume-index=1" <hostname>:/<volname> <mountpoint>Thanks for that tip! Just a question: as our clients can migrate beetween buildings (not often, but it happens), does using something like 'mount -o "remount,xlator-option=*replicate*.read-subvolume-index=X" …' will allow the client to point to the new target without stopping volume availability for clients? -- Y.-Krutika ------------------------------------------------------------------------ *From: *"Yannick Perret" <[email protected]> <mailto:[email protected]> *To: *"Krutika Dhananjay" <[email protected]> <mailto:[email protected]> *Cc: *[email protected] <mailto:[email protected]> *Sent: *Friday, March 4, 2016 3:35:57 PM *Subject: *Re: [Gluster-users] Per-client prefered server?
smime.p7s
Description: Signature cryptographique S/MIME
_______________________________________________ Gluster-users mailing list [email protected] http://www.gluster.org/mailman/listinfo/gluster-users
