I recently noticed that snapvault is failing because there is not enough space on the primary filer volume, and thus snapshots are automatically removed to free up space (otherwhise the lun that is in this volume will probably run full with the lun failing).
To prevent snapvault from failing I want to make the volume on the primary bigger.
The volume is on aggr3_fc
delng153> df -Ah
Aggregate total used avail capacity
aggr3_fc 4675GB 4104GB 571GB 88%
aggr3_fc/.snapshot 246GB 231GB 14GB 94%
So there seems to be plenty of space to resize the volume.
However what I wonder if there is any recommended percentage of space I should keep free on the aggregate to keep the filer fully functional (as in both functionality and performance).
Furthermore I wonder if I can size the volume down later if necessary without any interuptions or if this is only possible in very specific cases.
I mean as in free space, as long as the Flexvol is not using the space the volume can be resized.
It is alot, but depending on who you talk to at Netapp, depends on how the 80-85% is calculated.... Some say that it is the figure that you see on the aggregate utilisation and some say that it is the overall free space, including free space in the contained volumes. I think it also depends on whether or not you have thin provisioned your volumes.
I can see from the output that you have an aggregate snapshot reserve configured, by default 5% of the aggregate is set aside in the aggregate snapshot reserve. Unless you are are using syncromous mirroring or plexing your aggregate this 5% is wasted space, I wouldn't remove it completely as it has been a life line for me in the past, but 5% could be reduced to 3%. I would also remove the schedule for the aggregate snapshots. Use the following commands ("snap reserve -A aggr3_fc 3" "snap sched -A aggr0 0 0 0"). To answer your question..... by lowering the snapshot reserve, it should give you back roughly 100g.... you could potentially use this space to increase the volume and be no worse of then you were before....
As a rule of thumb I always look to keep 80% free space in the aggregate, but am less concerned about the free space in the volume. Therefore I thin provision (guarntee=none) the volume and then just manage the aggregate free space accordingly.
Personally I leave the aggregate snapshot reserve at the defaut 5% but do change the snap sched to 0 1 0, just to give myself a life line should I need to restore a volume.
It is recommended to keep a minimum of 3 % of space kept free on the aggregate . Which means that the aggregate should not be used more that 97% ever.
Personally I would recomend the aggregate is not used more that 95% . The extra 2% will come in handy if all your aggregate / volume space is used .
Yes you can resize the volume down later if necessary without any interuptions provided there is space unused in the volume. For example you have a 100G volume which has used 60G already . You can now resize the volume down upto 60G , Please note that the volume used space becomes 100% now and run out of space. so you may want to resize to 70G or 80G depending upon your need to get the space back to aggregate .
Netapp recomends to grow ( add disks) your aggregate when the utilization approach 80% - 85 % for better performance.
Thanks for your detailed answer! I'll surely keep this in mind in the future!
I just checked on the SAN and found out that the highest aggregate usage is 90%. I might be able to free up some space with some larger volumes. Adding new disks(disk cabinet) is not an option at the moment. Especially as the LUNs in the volumes still have alot of free space.