[Gluster-users] gluster replacing a brick

[Hariharan Thantry]

Shyam,

Thanks. What would be *extremely* useful is the ability to move data off
the brick just through replace brick, where the other identified brick is a
member of the current cluster (Isn't that something that would be natural?)

The *real* issue for me is that my disk hosts multiple bricks, each of
whose replicas are on other bricks (higher redundancy). You can't "remove"
a brick without taking out its replica as well. Wouldn't it be much simpler
for Gluster to be able to manage this in the backend through replace-brick?

BTW, I also noticed that Gluster doesn't like it if I went down from X
replicas to 1 replicas, and the X replica bricks are not specified in the
order as obtained from `gluster volume info`.

Thanks,
Hari


On Thu, Jan 30, 2014 at 8:24 PM, Shyamsundar Ranganathan <
srangana at redhat.com> wrote:

> > From: "Hariharan Thantry" <thantry at gmail.com>
> > To: gluster-users at gluster.org
> > Sent: Friday, January 31, 2014 4:23:03 AM
> > Subject: [Gluster-users] gluster replacing a brick
>
> > Hi folks,
> > Is there an easy way to shrink the size of a gluster volume by replacing
> some
> > of its constituent disks, and moving the data to other constituent disks?
>
> Wouldn't remove-brick suffice for what you have in mind?
> I would think you need to remove-bricks and rebalance to get the data
> migrated out of the removed bricks, and then decommission those bricks
> (commit the remove-brick).
>
> > I did
> > gluster volume replace-brick <volname> <old-brick> <new-brick>
> > where both <old-brick> and <new-brick> are part of the gluster setup,
> but it
> > wouldn't let me do this complaining that:
> > Brick: <new-brick> not available. Brick may be containing or be
> contained by
> > an existing brick.
>
> So <new-brick> is part of the same gluster volume, and hence the error
> (even if it was part of another gluster volume the same would happen)
>
> > Can I remove the new-brick, wipe off the gluster artifacts [the
> trusted.XXX
> > parameters], and then do the replace-brick? Will that cause any data
> loss?
>
> <new-brick> is part of the gluster volume, so wiping off gluster artefacts
> would in turn result in data loss, or rather gluster failing to see this
> brick as a part of the volume etc.
>
> > Thanks,
> > Hari
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://supercolony.gluster.org/pipermail/gluster-users/attachments/20140130/b0fc2e35/attachment.html>