Re: Add single server

[Joe Julian <joe@xxxxxxxxxxxxxxxx>]

On 05/01/2017 11:47 AM, Pranith Kumar Karampuri wrote:

On Tue, May 2, 2017 at 12:14 AM, Shyam <srangana@xxxxxxxxxx> wrote:

On 05/01/2017 02:42 PM, Pranith Kumar Karampuri wrote:

On Tue, May 2, 2017 at 12:07 AM, Shyam <srangana@xxxxxxxxxx
<mailto:srangana@xxxxxxxxxx>> wrote:

    On 05/01/2017 02:23 PM, Pranith Kumar Karampuri wrote:



        On Mon, May 1, 2017 at 11:43 PM, Shyam <srangana@xxxxxxxxxx
        <mailto:srangana@xxxxxxxxxx>

        <mailto:srangana@xxxxxxxxxx <mailto:srangana@xxxxxxxxxx>>> wrote:

            On 05/01/2017 02:00 PM, Pranith Kumar Karampuri wrote:

                    Splitting the bricks need not be a post factum
        decision, we can
                    start with larger brick counts, on a given node/disk
        count, and
                    hence spread these bricks to newer nodes/bricks as
        they are
                added.


                Let's say we have 1 disk, we format it with say XFS and that
                becomes a
                brick at the moment. Just curious, what will be the
        relationship
                between
                brick to disk in this case(If we leave out LVM for this
        example)?


            I would assume the relation is brick to provided FS
        directory (not
            brick to disk, we do not control that at the moment, other than
            providing best practices around the same).


        Hmmm... as per my understanding, if we do this then 'df' I guess
        will
        report wrong values? available-size/free-size etc will be
        counted more
        than once?


    This is true even today, if anyone uses 2 bricks from the same mount.


That is the reason why documentation is the way it is as far as I can
remember.



    I forgot a converse though, we could take a disk and partition it
    (LVM thinp volumes) and use each of those partitions as bricks,
    avoiding the problem of df double counting. Further thinp will help
    us expand available space to other bricks on the same disk, as we
    destroy older bricks or create new ones to accommodate the moving
    pieces (needs more careful thought though, but for sure is a
    nightmare without thinp).

    I am not so much a fan of large number of thinp partitions, so as
    long as that is reasonably in control, we can possibly still use it.
    The big advantage though is, we nuke a thinp volume when the brick
    that uses that partition, moves out of that disk, and we get the
    space back, rather than having or to something akin to rm -rf on the
    backend to reclaim space.


Other way to achieve the same is to leverage the quota functionality of
counting how much size is used under a directory.

Yes, I think this is the direction to solve the 2 bricks on a single FS as well. Also, IMO, the weight of accounting at each directory level that quota brings in seems/is heavyweight to solve just *this* problem.

I saw some github issues where Sanoj is exploring XFS-quota integration. Project Quota ideas which are a bit less heavy would be nice too. Actually all these issues are very much interlinked.

It all seems to point that we basically need to increase granularity of brick and solve problems that come up as we go along.

I'd stay away from anything that requires a specific filesystem backend. Alternative brick filesystems are way too popular to add a hard requirement.

 

            Today, gluster takes in a directory on host as a brick, and
        assuming
            we retain that, we would need to split this into multiple
        sub-dirs
            and use each sub-dir as a brick internally.

            All these sub-dirs thus created are part of the same volume
        (due to
            our current snapshot mapping requirements).




        --
        Pranith




--
Pranith

--

Pranith

_______________________________________________
Gluster-users mailing list
Gluster-users@xxxxxxxxxxx
http://lists.gluster.org/mailman/listinfo/gluster-users

_______________________________________________
Gluster-users mailing list
Gluster-users@xxxxxxxxxxx
http://lists.gluster.org/mailman/listinfo/gluster-users