The NOW database suggests that bug 2689 ("sysstat output is inaccurate if displayed at a long interval") is fixed in recent versions of ONTAP, e.g. 6.1.1.

I don't think so.

See the start of the disk scrub tonight on an F740...

["sysstat -u 60" is running]

CPU Total Net kB/s Disk kB/s Tape kB/s Cache Cache CP CP Disk ops/s in out read write read write age hit time ty util 1% 21 14 84 35 87 0 0 >60 100% 2% T 1% 1% 7 8 11 38 88 0 0 >60 100% 2% T 2% Sun Sep 9 01:00:00 BST [kern.uptime.filer:info]: 1:00am up 43 days, 6:54 194462270 NFS ops, 0 CIFS ops, 0 HTTP ops 16% 17 57 126 10546 141 0 0 >60 100% 3% T 18% 54% 40 572 520 1887065754 824 0 0 >60 100% 11% 5 62% 49% 13 44 56 3537004865 135 0 0 >60 99% 5% T 57% 51% 30 40 233 547830010 129 0 0 >60 99% 5% T 59%

The unreasonable reported disk read rates are apparently due to more than 2^31 bytes being read during the recording interval. "sysstat -u 45" gets the numbers back in range:

CPU Total Net kB/s Disk kB/s Tape kB/s Cache Cache CP CP Disk ops/s in out read write read write age hit time ty util 50% 9 15 65 37010 91 0 0 >60 100% 4% T 59% 51% 18 87 100 37068 188 0 0 >60 100% 4% T 59% 50% 5 10 9 37586 78 0 0 >60 100% 3% T 59% 51% 28 218 247 36076 327 0 0 >60 100% 5% T 58%

Chris Thompson University of Cambridge Computing Service, Email: cet1@ucs.cam.ac.uk New Museums Site, Cambridge CB2 3QG, Phone: +44 1223 334715 United Kingdom.