Space reserved LUN offline because volume full

[THIERRY_D]

Hi All,

Got a question that is bugging me. As I understand a LUN can't be taken offline if volume is guaranteed and LUN space reservation is enabled ? Right ?

Here is the configuration for the LUN:

	/vol/v_logs/q_logs/logs.lun   90.0g (96638814720)   (r/w, online, mapped)
	Comment: "Lun logs"
	Serial#: dem5GZfo9afr
	Share: none
	Space Reservation: enabled
	Multiprotocol Type: windows_gpt
	Maps: viaRPC.xxxxxxxx
	Occupied Size:	5.4g (5753962496)
	Creation Time: Fri Oct 14 15:40:32 CEST 2011
	Cluster Shared Volume Information: 0x0

Here for the volume :

nosnap=on, nosnapdir=off, minra=off, no_atime_update=off, nvfail=off,

ignore_inconsistent=off, snapmirrored=off, create_ucode=on,

convert_ucode=on, maxdirsize=327680, schedsnapname=ordinal,

fs_size_fixed=off, guarantee=volume, svo_enable=off, svo_checksum=off,

svo_allow_rman=off, svo_reject_errors=off, no_i2p=off,

fractional_reserve=50, extent=off, try_first=volume_grow,

read_realloc=off, snapshot_clone_dependency=off, dlog_hole_reserve=off,

nbu_archival_snap=off

Here the logs saying the LUN has been taken offline :

Wed Dec  5 14:38:41 CET [FILER01:wafl.vol.full:notice]: Insufficient space on volume v_logs to perform operation. 60.0KB was requested but only 4.00KB was available.

Wed Dec  5 14:38:41 CET [FILER01:wafl.write.fail.spcres:warning]: Write failed to file with space reservations due to lack of disk space in volume v_logs (guarantee volume, inode 98, offset 13609274368, len 61440).

Wed Dec  5 14:38:41 CET [FILER01:callhome.tgt.lun.nospc:CRITICAL]: Call home for LUN OUT OF SPACE

Wed Dec  5 14:38:41 CET [FILER01:scsitarget.lun.noSpace:error]: LUN '/vol/v_logs/q_logs/logs.lun' has run out of space.

Wed Dec  5 14:38:41 CET [FILER01:lun.offline:warning]: LUN /vol/v_logs/q_logs/logs.lun has been taken offline

So thats the thing, am I wrong to say that if a LUN is space reserved and the volume size guaranteed, the LUN should not be taken offline even if the volume is full ?

I imagined , snapshot would fail and all but the lun would still be online ?

Is there something I did not see ?

Any idea would be welcomed

Thierry

[aborzenkov]

Only if fractional reserve is set to 100%. In your case it is set to 50%.

[THIERRY_D]

okay but 50% of 90Gb is still 45Gb that were reserved for possible changes , so the 60KB necessary should have been available.

I imagine it would have failed if I would have needed to write more than 45Gb ?

[aborzenkov]

Please show output of “df -r”

[THIERRY_D]

Filesystem	kbytes   used   avail	reserved	Mounted on
/vol/v_logs/	204472320  196683600 7788720	2809564	/vol/v_logs/
/vol/v_logs/.snapshot	0   99054868      0	0	/vol/v_logs/.snapshot