ONTAP Discussions
ONTAP Discussions
Hello folks,
I establish a snapmirror relation ship between a source volume of size 3.0TB to destination volume 3.5TB.
when I initialize the snapmirror process it says snapmirror aborted.
Then i started creating snapshot manually but it says no free disk space on source volume I already started increasing the size of the source volume but still it tell's me no free disknspace.
please solve my issue as i have 3 lun's are carved on source volume and it is in production.
Thanks in advance.
Naveen
Solved! See The Solution
Hi Naveen,
If you have a LUN inside your volume and use snapshots it would be prudent to set volume autosize on. Look up the command
and see what it does. Fractional reserve is a mechanism that allows to take space upfront for the LUN so that it never runs out of space.
this has been turned off now as you know, so now you need to allow the volume to grow in case it needs to. Do you use snapshots in the volume?
Eric
Naveen,
Can you look in "/etc/log" for snapmirror log and find an error message at the time you try to snapmirror intitialize. I assume you have not
had this working and initializing is failing, is this correct?
then please send us this, please note from wich controller I need the info, please only send for the volume in question>
source:
df -g
vol options
snap list
destination
df -g
vol options
snapmirror status -l <volume_name>
snap list
df -r
Thanks,
Eric
Eric,
My volume name is EV on source filer & destination filer S_EV.
source filer details:
snap list EV
working..
no snapshots exists.
> vol options EV
nosnap=off, nosnapdir=off, minra=off, no_atime_update=off, nvfail=off,
ignore_inconsistent=off, snapmirrored=off, create_ucode=on,
convert_ucode=on, maxdirsize=31457, 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=100, extent=off, try_first=volume_grow
df -g:
Filesystem total used avail capacity Mounted on
/vol/EV/ 4183GB 3422GB 760GB 82% /vol/EV/
/vol/EV/.snapshot 0GB 0GB 0GB ---% /vol/EV/.snapshot
Destination details:
df -g
I cannot see any thing about S_EV volume
snapmirror status -l S_EV
snapmirror status -l S_EV
Snapmirror is on.
Source: BOSNAP002:EV
Destination: BOSNAP006:S_EV
Status: uninitialized
Progress: 0KB
State: Uninitialized
Lag: -
Mirror Timestamp: -
Base Snapshot: -
Current Transfer Type: Initialize
Current Transfer Error: -
Contents: -
Last Transfer Type: -
Last Transfer Size: -
Last Transfer Duration: -
Last Transfer From: -
df -r
no information about S_EV
when I initialize snapmirror I get "snapmirror process aborted"
please correct me.
Thanks in Advance
naveen.
hello eric,
on destination vol options S_EV
volume S_EV resticted or offline options are unknown.
sorry for missing it on earlier reply.
Thanks in advance
Naveen
Hi,
I noticed your source volume is a lot bigger than 3.0TB, its more like 4.5TB. Your destination should be the same size as a minimum.
For starters you need to online the destination volume
vol online S_EV then you need to resize it, but before you do check you have enough size in the aggregate:
df -Ag look for the aggr. where S_EV is, and see that you have enough space. If you do, then run
vol size S_EV 4500g at this point you could also get a vol options and send across
vol options S_EV Restrict the volume again
vol restrict S_EV Snapmirror initialize:
snapmirror initialize -S BOSNAP002:EV BOSNAP006:S_EV check status
snapmirror status -l
You should be OK at this stage. If you find that you have not got enough space you can shrink destination volume while it is online and serving
data and set the destination volume the same size. Keep in mind a volume needs to be online to have the size changed or have commands ran against it.
Also, make sure you have a /etc/snapmirror.conf file with a schedule in it.
Eric
Hello eric,
I increased size of the destination (S_EV) volume bigger than source volume.
Again I intialized the snapmirror again I got a error "snapmirror process was aborted"
My concern since i created LUN;s on my source volume EV do i need to make changes on
vol options EV
nosnap=off, nosnapdir=off, minra=off, no_atime_update=off, nvfail=off,
ignore_inconsistent=off, snapmirrored=off, create_ucode=on,
convert_ucode=on, maxdirsize=31457, 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=100, extent=off, try_first=volume_grow
do u want me to change "fractional_reserve=0" instead of 100.
Thanks
Naveen.
ional reserve @ 0Hi,
Your VOL are SIS or not ?
If no , you can effectivly try to made a fracttional reserve @ 0, beceause when your snapmirror is launch the first think it's do, is a snapshot, and with the Fract reserve @ 100, the snap want to take the same space of the original, if I remembre correctly, my courses ...
Anyone confirm ?
And you dont have any space to do that
So if you turn to 0 your fract reserve, be careful ! the snap can take space on vol, and your lun can have no space to write on VOL...
It's highly recommended to check very frequently the space avaiable ...
Hi,
Can you run
df -r
on both volumes and post it here. Also, go into /etc/log/ and look for snapmirror log and get the error message that relates to this volume.
For now leave fract.reserve to what it is. Turning that on or off is a bigger decision about how you monitor space and how you use snapshots on the volume.
We use the following for our LUNs
no snap reserve
no volume guarantee
no space reservation for the LUN
this pushes all the free blocks into the aggregate and we monitor space utilization at aggr. level. that avoids us having to keep track of snap reserver, volume
space used. its simplifies everything and enables management to keep look at one graph to decide when to buy storage.
Eric
Hello eric,cedric,
Thanks for your support I changed fractinaol_reserve to 0% and started the snapmirror this time it got successful.
following is the message:
vol options EV
nosnap=off, nosnapdir=off, minra=off, no_atime_update=off, nvfail=off,
ignore_inconsistent=off, snapmirrored=off, create_ucode=on,
convert_ucode=on, maxdirsize=31457, 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=0, extent=off, try_first=volume_grow
df -r EV
df -r EV
Filesystem kbytes used avail reserved Mounted on
/vol/EV/ 4386193408 3588930688 797262720 0 /vol/EV/
/vol/EV/.snapshot 0 40644 0 0 /vol/EV/.snapshot
destination:
df -r S_EV
df -r S_EV
Filesystem kbytes used avail reserved Mounted on
/vol/S_EV/ 4386193408 3588909680 797283728 0 /vol/S_EV/
/vol/S_EV/.snapshot 0 19636 0 0 /vol/S_EV/.snapshot
> snapmirror status -l S_EV
Snapmirror is on.
Source: BOSNAP002:EV
Destination: BOSNAP006:S_EV
Status: Idle
Progress: -
State: Snapmirrored
Lag: 00:53:20
Mirror Timestamp: Wed Oct 14 08:01:23 EST 2009
Base Snapshot: BOSNAP006(0101184528)_S_EV.40
Current Transfer Type: -
Current Transfer Error: -
Contents: Replica
Last Transfer Type: Scheduled
Last Transfer Size: 19644 KB
Last Transfer Duration: 00:02:32
Last Transfer From: BOSNAP002:EV
If any thing needsto be corrected please let me know
Thanks
Naveen
Hi Naveen,
If you have a LUN inside your volume and use snapshots it would be prudent to set volume autosize on. Look up the command
and see what it does. Fractional reserve is a mechanism that allows to take space upfront for the LUN so that it never runs out of space.
this has been turned off now as you know, so now you need to allow the volume to grow in case it needs to. Do you use snapshots in the volume?
Eric
yes eric,
I have a LUN inside the volume we monitor the volume & aggregate and delete the old snapshots in the volume,.
Thanks
Naveen.
Hi Naveen,
So now you monitor volume AND aggregate. If you were to set volume guarantee to none you would only have to monitor aggr. space.
This is thin provisioning. Of course you need to make sure you never fill up your aggr. above 90%, or you have to move things around
and/or buy more disk space. Having snap autodelete should also help, I assume you already got snap autodelete on.
Eric
yes Blair, we have snap auto delete option on.
Thanks for you help.
Thanks
Naveen.