I have a script that performs backups for an
application. Part of this script is a snapvault
of volumes with just Luns. After the snapvault
finishes, I run a “snap delta” on the snapvault
target vol to determine the time span between
the most recent and oldest snapshots/backups.
This is check to make sure snapvaults are working.
Sometimes after a lot of changes on the lun, and thus
a large snapvault operation, the snap delta can run for hours!!!
All filers are:
IBM N2857 (NetApp FAS2240)
Data ONTAP Release 8.1.2P1 7-Mode
For example, here is the snapstatus of a snapvault after it
completed. It took 17m and transferred 6.4gb. The qtree
contains about 5 luns totaling to about 450GB.
Source: xxxx:/vol/v_asdf_db/q_asdf_db
Destination: yyyy:/vol/v_asdf_db/q_asdf_db
Status: Idle
Progress: -
State: Snapvaulted
Lag: 00:20:10
Mirror Timestamp: Tue Dec 24 16:39:33 EST 2013
Base Snapshot: yyyy(1896301030)_v_asdf_db-base.0
Current Transfer Type: -
Current Transfer Error: -
Contents: Replica
Last Transfer Type: Update
Last Transfer Size: 6450652 KB
Last Transfer Duration: 00:16:59
Last Transfer From: xxxx:/vol/v_asdf_db/q_asdf_db
After the above snapvault, the script ran a “snap delta”
on vol xxxx:/vol/v_asdf_db (snapvault target).
The cmd took +2hr to run!!!
(from my script – note the start and end time stamps in form yyyymmddhhmmss)
====> CheckSnapAge - Start 20131224171935
==> cmd = rsh yyyy -n snap delta v_asdf_db v_asdf_db.31
v_asdf_db.31 Active File System 989985036 9d 17:51 4233243.016
====> CheckSnapAge - End 20131224193203
In this case, the snapvault target spans 9 days of
backups. To determine this took +2 HOURS!
The snapvault target isn’t getting hit very hard at all,
so resources don’t appear to be the limitation.
I’ve looked for other ways to determine
The time span of the snaps, but haven’t come up with any.
{The dates on a “snap list” don’t have the year!)
Q) Why can a “snap delta” run a very long time (but not always)?
Thanks for any thoughts!
Rick