This is idd normal behaviour. When blocks are getting filled up in your lun, Netapp will also notice that and the used space will grow. When however your file system frees blocks, this is not told to Netapp, so those blocks will be kept in use even when there is no data in it anymore. Netapp simply doesn't know they are freed.
 
In Windows you can use Snapdrive 5 with a new feature called 'Space reclamation". In this case Snapdrive will scan your NTFS filesystem regularily and will tell Netapp which blocks can be freed. In this case your used space will also decrease on Netapp when you free up space in your lun.
 
Grtz,
Tom

From: owner-toasters@mathworks.com [mailto:owner-toasters@mathworks.com] On Behalf Of Daniel Keisling
Sent: vrijdag 25 januari 2008 23:15
To: toasters@mathworks.com
Subject: RE: Thin provisioning LUNs, A-SIS, and notifications

After a couple of replies, I have determined that the volume never decreasing is indeed normal behavior.  I started SIS on the volume and the LUN did decrease, with great results.
 
Now if I can just configure SNMP notification for volume autosize events....


From: owner-toasters@mathworks.com [mailto:owner-toasters@mathworks.com] On Behalf Of Daniel Keisling
Sent: Friday, January 25, 2008 12:07 PM
To: toasters@mathworks.com
Subject: Thin provisioning LUNs, A-SIS, and notifications

Hello,
 
I am starting to thin provision LUNs (lun set reservation disable) in order to start using A-SIS on some large (1TB) volumes.  Before I turn A-SIS on, I started testing these LUNs and do see the space on the filer grow accordingly with writes on the host.  However, the space never seems to go down when I delete files on the host:
 
rtpstore2b> df /vol/iscsi
Filesystem              kbytes       used      avail capacity  Mounted on
/vol/iscsi/           10485760    9974748     511012      95%  /vol/iscsi/
/vol/iscsi/.snapshot          0          0          0     ---%  /vol/iscsi/.snapshot
 
 
[root@vrhel5 /]# df /mnt/iscsi
Filesystem           1K-blocks      Used Available Use% Mounted on
/dev/sde1             10321192   7804844   1992064  80% /mnt/iscsi
 
 
I see that when I begin writing again, the filer space won't start increasing until there is an increase in space above what the filter already thinks.  Is this behavior normal?  When I turn SIS on, will the filer LUN size reduce as blocks are deduplicated so that I can then reduce the volume?
 
Lastly, I will definitely need to know when my volumes autosize so I can focus my attention on not having the aggregate fill up.  It would be wonderful if I could set up an SNMP trap to send to Nagios for notification.  Before I dig into the MIBs and create the custom trap (if I even can), has anyone done anything similar they could share?
 
TIA,

Daniel
 
 
 
 




______________________________________________________________________
This email transmission and any documents, files or previous email
messages attached to it may contain information that is confidential or
legally privileged. If you are not the intended recipient or a person
responsible for delivering this transmission to the intended recipient,
you are hereby notified that you must not read this transmission and
that any disclosure, copying, printing, distribution or use of this
transmission is strictly prohibited. If you have received this transmission
in error, please immediately notify the sender by telephone or return email
and delete the original transmission and its attachments without reading
or saving in any manner.




______________________________________________________________________
This email transmission and any documents, files or previous email
messages attached to it may contain information that is confidential or
legally privileged. If you are not the intended recipient or a person
responsible for delivering this transmission to the intended recipient,
you are hereby notified that you must not read this transmission and
that any disclosure, copying, printing, distribution or use of this
transmission is strictly prohibited. If you have received this transmission
in error, please immediately notify the sender by telephone or return email
and delete the original transmission and its attachments without reading
or saving in any manner.