today I think we hit hard this bug, filling a 3.7TB aggregate with 1.2TB of A-SIS
Total space WAFL reserve Snap reserve Usable space
BSR NVLOG A-SIS Smtape
3673GB 367GB 0KB 3306GB
0KB 1122GB 0KB
Space allocated to volumes in the aggregate
Volume Allocated Used Guarantee
volroot 163GB 4167MB volume
linux_virtual_machines 905GB 900GB none
test_iscsi_vol 4487MB 4308MB none
microsoft_virtual_machines 106GB 105GB none
iscsi_linux_vms_vol 138GB 137GB none
mail_fast 707GB 700GB none
Aggregate Allocated Used Avail
Total space 2027GB 1853GB 156GB
Snap reserve 0KB 0KB 0KB
WAFL reserve 367GB 39GB 327GB
Now we cleaned up a little bit and now we have 150GB free and the services running.
I would like to resolve the issue, and I'm asking myself if it's wise to:
- Disable the SIS avoiding a tonight disk filling:
sis off /vol/<volname>
for every volume that has the deduplication on
- When convenient (maybe in the weekend) run
sis start -s /vol/<volname>
- When convenient ask my specialist to upgrade the DataONTAP to a fixed version
Am I right?
PS: I'm also asking myself if there is a way to see how many stale fingerprints I have.
the percentages shown in sis status -l are not too high and it seems to me that they are wrong!