2014-05-16 08:11 AM
Ontap Version 8.1.3P1
Root aggr is 2 disk raid4 aggr with root vol on 7k Sata.
Every hour the root vol iops spike to 20k - 22k and I have no idea why. During this time the root vol latency stays low, however there is a performance impact on the other data aggr on the system.
I have used dfm to view the spikes and see that the ops are all sys_ops.
There is no SIS running and this filer is the snapmirror destination for only one volume.
The filer is used to serve cifs and nfs data only.
Reallocation scan are on for 4 volumes in the data aggr:
Interval: 1 day
There are no other scheduled jobs on this filer that I know of.
Could the reallocation be the cause for the huge spike in iops? Has any one seen anything like this before?
2014-05-19 11:55 AM
I verified that there is only one hostname in the snmp.access list. The interval of the spike does seem to point to some sort of system query such as from snmp or perhaps the dfm server that is monitoring it.