[Gluster-devel] Re:The QUOTA xlator V0.2 codename "dual fire"

Angel clist at uah.es
Tue Jan 22 00:07:23 UTC 2008


Hi

WARNING: quota is still under development,is not production quality nor yet accepted by gluster team !! 

/* do no use on pord systems yet  */

Use as this:

Currently QUOTA acts as overal 'limit' to all clients and users accesing the volume.
 
volume brick
  type storage/posix                   # POSIX FS translator
  option directory /home/export        # Export this directory
end-volume

### 'Quota' feature should be added on the server side (with posix volume as subvolume) where it can be used to limit max usage,
on the underlaying block device.
volume quota
  type features/quota
  subvolumes brick
  option max-size 50M   # Limit underlaying volume to 50MB of space
  option enforce-level  # Limit upper schedulers or limit all oprations (CURRENTLY NOT IMPLEMENTED)
end-volume

volume server
  type protocol/server
  subvolumes quota brick
  option transport-type tcp/server
  option auth.ip.brick.allow 192.168.* # Allow access to "brick" volume
  option auth.ip.quota.allow 192.168.* # Allow access to "quota" volume (exporting 50MB disk space)
end-volume

// Any client will see that volume quota has only 50Mb available whereas brick0 its only constrained to real storage available blocks.

So any module on client will just use as much as 50Mb of space from the sever where the directory os poisx-brick0 resides.
If you make a df on client you will see Total: 50MB Used:0 Free: 50MB.

If you need quotas on user, use quota facilities of native filesystemn on server..

Regards, Angel


El Martes, 22 de Enero de 2008 matthew zeier escribió:
> Is this a per-directory quote or per gluster mount?
> 
> (I need the former)
> 
> Angel wrote:
> > Hi all
> > 
> > Im proudly present my QUOTA xlator v0.2 'dual fire'.
> > 
> > "The Quota translator imposes a limit on the underlaying storage allowing to limit disk usage 
> > of subvolumes directly on filesystem calls or by means of informing schedulers on stats mops."
> > 
> > I like to thank avati for his great help on getting me get the inners of glusterFS.
> > 
> > This is the present features (V0.2):
> > 
> >         - Check limits and reports limited values from stats mops to upper schedulers (unify et al).
> >         - Cheks limits and reports limited values on statsvfs fops (for "df " command)
> >         - Check limits on writes, reporting ENOSPC on hitting max-size writes
> >         - Checks on unliks computing recovered space.
> > 		- Also checks for multiple hardlinks and account on last hardlink unlinked.
> > 
> > Changelog:
> > 	Changed unlink tracking scheme "stat before and unlink after" that fired unlink on stat_cbk
> > 	on pipeline firing both calls for performance gains.
> > 	Changed passing struct stat *buf from posix-storage across call to a private copy on quota xlator  
> > 
> > 
> > Include docs at /doc/examples/quota.vol
> > 
> > Comments are welcome.
> > 
> > Keep up the Good Work(tm) GlusterFS guys!!!
> > 
> > Regards, Angel
> > 
> > 
> > ------------------------------------------------------------------------
> > 
> > _______________________________________________
> > Gluster-devel mailing list
> > Gluster-devel at nongnu.org
> > http://lists.nongnu.org/mailman/listinfo/gluster-devel
> 



-- 
------------------------------------------------
Clist UAH
------------------------------------------------

-------------------------------------------------------

-- 
------------------------------------------------
Clist UAH
------------------------------------------------





More information about the Gluster-devel mailing list