Hi Jeff,
There is a thread on the RAID lost write scrub when upgrading to Ontap 8.1 or later:
https://forums.netapp.com/thread/30084
It's also described in the Netapp KB ID: 3013583 "What is RAID lost write protection?" This mentions:
"The only performance impact expected is that of scrub, which can be scheduled and stopped by the usual means (for more information, see the 'aggr scrub' man pages)."
You can delay this RLW upgrade by tuning the scrub policies to periods of lower usage and in your case stagger them across the aggregates.
I can't find references to it but even prior to 8.1 I noticed for a time after an Ontap upgrade there is an increase in disk and CPU activity while the Filer does background WAFL tasks post upgrade.
As the Netapp Upgrade Advisors suggest I always run a perf stat before an Ontap upgrade during a period when the Filer is under it's normal load so that I have something to compare with post upgrade.
If you are considering 8.1.2 I would take a look through the bugs fixed in the subsequent P releases (8.1.2P4 has just been released) as there are quite a few important fixes in these.
On a side note if you have multiprocessor Controllers the distribution of load across CPUs was improved dramatically from 7.x.x to 8.1 and above which should improve CPU utilization.
Regards Martin
-- View this message in context: http://network-appliance-toasters.10978.n7.nabble.com/Delaying-64-bit-aggreg... Sent from the Network Appliance - Toasters mailing list archive at Nabble.com.