2013-06-18 04:48 PM
Here's the short version:
Q: Is it possible to have a situation where a volume can be physically larger than its auto-grow maximum?
Here's the long version:
We have a tool we've written here in house that checks two things -- First, to see if all of our volumes have auto-grow enabled, and second, whether or not the auto-grow maximum is set sufficiently high that we don't at some point face a situation where we have a volume or LUN go offline on us unexpectedly.
This tool has been performing nicely...but the other morning, we noticed something unusual in the syslog output of one of our filers; that ONTAP was having trouble auto-growing the size of a particular volume. Here's the cut 'n paste of what we see (Note: replaced the actual volume name for 'foo', for employer confidentiality)
wafl.vol.autoSize.fail:info]: Unable to grow volume 'foo' to recover space: Volume cannot be grown beyond maximum growth limit #015.
So, we had a look at the volume from the command line. Here's what we saw:
>vol size foo
vol size: Flexible volume 'foo' has size 5t.
>df -g foo
Filesystem total used avail capacity Mounted on
/vol/foo/ 5120GB 912GB 4207GB 18% /vol/foo/
/vol/foo/.snapshot 0GB 0GB 0GB ---% /vol/foo/.snapshot
>vol autosize foo
Volume autosize is currently ON for volume 'foo'.
The volume is set to grow to a maximum of 3000 GB, in increments of 63 GB.
So, long story short, it looks like (to us) that it's possible to have a volume that had, at one point, an auto-grow maximum set...but was later exceeded by _manually_ increasing its size. Is this possible/legal in ONTAP? Or is the notion of an auto-grow maximum and a volume size maximum two independent ideas?