[Gluster-devel] online resizing of backend storage

[Anand Avati]

August,
 resizing this way is safe. posix export's size is not 'used' anywhere. The
only place I can see is the ALU scheduler of unify to decide where to
schedule the next create. But the ALU scheduler frequently updates the stats
and should pick the new size in the next probe.

avati

2007/8/1, August R. Wohlt <glusterfs at isidore.net>:
>
> Hi glusterfs hackers,
>
> In my server setup, I am exporting ext3 volumes as posix bricks to my
> glusterfs clients. The ext3 volumes I am exporting are living on top of
> LVM
> logical volumes. I like to grow these logical volumes and their
> corresponding ext3 file system often. So one day if I do a df on the
> client,
> I might see a 200G volument, but the next day this is grown to 300G. I
> have
> never restarted the client or server after doing this, and so far I
> haven't
> seen any issues. When I am doing this, when I do a df on the client, it
> sees
> the new size and is happy.
>
> I would like to ask those of you more familiar with the code if there is
> any
> internal representation of the size of the server brick that might get
> confused if I resize the underlying file system, or is resizing quite
> alright (as long as i'm increasing, not decreasing the storage available
> to
> the server brick) without restarting anything.
>
> thanks!
> :goose
> _______________________________________________
> Gluster-devel mailing list
> Gluster-devel at nongnu.org
> http://lists.nongnu.org/mailman/listinfo/gluster-devel
>



-- 
Anand V. Avati