0

I get strange disk io nread (bytes read) counter values (from kstat) on Solaris. Example of collected nread value for sd6 disk collected in 30s interval (command kstat -n sd6):

768579416
768579416
768579416
768579416
768579416
768579416
768579416
768496080
768496080
768496080
768496080
768496080
768496080
768496080
768496080
768530896
768530896
768447560
768447560
768447560

One would suppose that the relative read bytes count can't be negative.. I wonder what can couse this situation and whether there is more reliable disk io data available.

Some info about the system:

machine:~ # uname -a
SunOS machine 5.10 Generic_127112-11 i86pc i386 i86pc
machine:~ # cat /etc/release 
                        Solaris 10 11/06 s10x_u3wos_10 X86
           Copyright 2006 Sun Microsystems, Inc.  All Rights Reserved.
                        Use is subject to license terms.
                           Assembled 14 November 2006
Improve this question
1
  • it happens on almost every machine, on each for several disks
    – mykhal
    Dec 21, 2009 at 10:46

1 Answer 1

Reset to default
0

Nreads should indeed grow, not decrease. What about the other statistics (reads, writes) and does the same behavior happen with other devices ? What Solaris release / kernel patch level are you using ? It might just be a bug.

Improve this answer
3
  • I have added the machine info to the question
    – mykhal
    Dec 21, 2009 at 11:17
  • What about the other statistics (reads, writes) ?
    – jlliagre
    Dec 24, 2009 at 0:10
  • other statistics are OK, not decreasing
    – mykhal
    Jan 1, 2010 at 6:03

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .