If it's VSM and FlexVols, it might be deswizzling.  I wouldn't worry much about it.  There are back-end WAFL
threads that run when nothing else runs.  If actual work shows up, they back off.  CPU isn't always a good
performance indicator.
 
Most of the other commands are hidden and not very customer-friendly.  Things like wafl_stat and statit
will show you all sorts of things, but you pretty much need to be trained on them to read them properly.
 

-- Adam Fox
NGS Tools Developer
adamfox@netapp.com

 


From: Steven Mandrake [mailto:srmanager@gmail.com]
Sent: Tuesday, April 03, 2007 4:25 PM
To: toasters@mathworks.com
Subject: Snapmirror NTAP is at 20% constantly

Hi Guys,

Any reason why CPU utilization on my snapmirror destination is always at 20%??? My production filer is much less than that. The Snapmirror only happens once per day. All other indicators seem idle. Here is a sample output:

CPU    NFS   CIFS   HTTP      Net kB/s     Disk kB/s      Tape kB/s    Cache

                               in   out     read  write    read write     age
 22%      0      0      0       1     1       65     59       0     0       4
 20%      0      0      0       1     0      121    120       0     0       4
 26%      0      0      0       1     1      336     59       0     0       3
 22%      0      0      0       1     1     2017    472       0     0       3
 23%      0      0      0       1     0     2461    413       0     0       3
 22%      0      0      0       1     0     2334    182       0     0       3
 24%      0      0      0       1     1     1307    206       0     0       3
 20%      0      0      0       1     1       61     59       0     0       3
 22%      0      0      0       1     0      123    118       0     0       3

Another question -- Is there a way I can monitor utilization/performance of NVRAM and Memory? What other perf commands can I use other than sysstat?

Thanks