FAS3040 Vol Full LUN Offline

[ZACH_DUNDORE]

Having some issues with my FAS3040.  The LUN I have created keeps going offline because there is not space left on the volume.  The wierd part about it is that my LUN is set to 4.8TB and my Volume is set to 4.83.  I have snapshots turned off, and the fractional reserve is set to 0.  What is consuming that last .03TB of my volume and causing my LUN to go offline?  This LUN is not spaced reserved but i didn't think i had to since i wasn't allocating more space than what was available on the volume.

Any help would be much appreciated.

Just to give you a complete picture, the LUN is connected to a Server 2008 R2 VM that is connected to the LUN via iSCSI and using it to store recorded IP video.  We really can't lose the data.

[ZACH_DUNDORE]

I did read this discussion (https://communities.netapp.com/message/80367#80367) but it doesn't appear to apply exactly because my Volume guarantee is set to volume.

[ZACH_DUNDORE]

As a side note, the filer view is reporting only 4.81TB used of the 4.83TB available on the volume:

Name:		vol1
Type:		Flexible		Root Volume?		-
FlexClone ?		-		Containing Aggregate:		aggr1
Status:		online,raid_dp
Used Capacity:		4.81 TB		Space Guarantee		volume
% Used:		100%		Language:		C
Total Capacity:		4.83 TB		Total Size:		4.83 TB

[ZACH_DUNDORE]

Another interesting bit is when connected to the device via serial cable, i run the command vol status -v and the status of the space guarantee of the volume says: guarantee=volume(disabled)

That appears to conflict with the status screen i see in the filer view.

[gestionesan]

Filer View is deprecated. You should use "onCommand" or better "OnCommand system manager". Try ... For the question, maybe you can apply an "autogrow" on the volume, in this way Your lun can't go offline. bye

[paul_wolf]

Autogrow in this case would immediately grow the FlexVol by the specified increment as it's above the 98% threshold (99.4% currently) for a FlexVol of that size. It will continue to grwo the FlexVol until the space utilization drops below 98%, the volume grows by 25% or the volume reaches the maximum allowable size which ever comes first.

[radek_kubka]

Hi,

Can you post the output of this command please?

> df -r vol1

Regards,

Radek

[paul_wolf]

Did you do a full format on the LUN? Is the LUN thin provisioned?

There isn't a lot of free space on the FlexVol and that free space is where changed blocks get written from the LUN. If the controller thinks that all 4.8 TB from the LUN is in use blocks, it can only write changes to the free space in the FlexVol and depending on the change rate in the LUN, .03 TB may not be enough space.

If you have the space, you really should increase the size of the FlexVol to a minimum of 10% larger than the LUN. Thin provision the FlexVol if you are that worried about space.

Can you post the output of the following:

vol options [volname]

df -h [volname]

lun show -v [ful LUN path]