Hi there,
Can anyone confirm my thoughts that the 'top' command only reports on
hard disk use WITHIN the box rather than anything attached to it
remotely (SAN or otherwise)?
Many thanks
Regards,
Hugo
|
|
0
|
|
|
|
Reply
|
Hugo
|
11/15/2006 7:58:06 PM |
|
On Wed, 15 Nov 2006 19:58:06 +0000, Hugo <hugo@nomorespamthanks.net> wrote:
> Hi there,
>
> Can anyone confirm my thoughts that the 'top' command only reports on
> hard disk use WITHIN the box rather than anything attached to it
> remotely (SAN or otherwise)?
top only shows CPU and memory use.
Something like iostat -x 5 will show disk (and NFS) activity).
A bientot
Paul
--
Paul Floyd http://paulf.free.fr (for what it's worth)
Surgery: ennobled Gerald.
|
|
0
|
|
|
|
Reply
|
Paul
|
11/15/2006 10:09:23 PM
|
|
Paul Floyd wrote:
> On Wed, 15 Nov 2006 19:58:06 +0000, Hugo <hugo@nomorespamthanks.net> wrote:
>> Hi there,
>>
>> Can anyone confirm my thoughts that the 'top' command only reports on
>> hard disk use WITHIN the box rather than anything attached to it
>> remotely (SAN or otherwise)?
>
> top only shows CPU and memory use.
>
> Something like iostat -x 5 will show disk (and NFS) activity).
What does the "IO Wait:" stand for in top's display? - I always thought
it was how many processes were "waiting" because of Disk Input/Output
activity?
Thanks again,
Hugo
|
|
0
|
|
|
|
Reply
|
Hugo
|
11/15/2006 11:24:14 PM
|
|
Hugo <hugo@nomorespamthanks.net> wrote:
> What does the "IO Wait:" stand for in top's display? - I always thought
> it was how many processes were "waiting" because of Disk Input/Output
> activity?
No, it's a CPU figure. It has nothing to do with # of processes. One
process (thread actually) doing I/O can drive the iowait to nearly 100
percent *if* the machine is otherwise idle.
The number of threads blocked on I/O is under the 'b' column in 'vmstat'
output.
Technically, the kernel maintains timestamps for when the CPU is running
user code, kernel/system code, and when the CPU is idle. I/O wait is a
subdivision of idle time.
The figure tends to be misinterpreted. It's not always an indication of
a problem, and it's not always present even when there is a problem. So
in isolation, it's not appropriate as an indicator of anything.
Due to the common misinterpretation, Solaris 10 has removed the
calculation. The value is forced to 0 and all CPU idle time is reported
strictly as 'idle'.
--
Darren Dunham ddunham@taos.com
Senior Technical Consultant TAOS http://www.taos.com/
Got some Dr Pepper? San Francisco, CA bay area
< This line left intentionally blank to confuse you. >
|
|
0
|
|
|
|
Reply
|
Darren
|
11/16/2006 12:17:37 AM
|
|
Hugo wrote:
> Hi there,
>
> Can anyone confirm my thoughts that the 'top' command only reports on
> hard disk use WITHIN the box rather than anything attached to it
> remotely (SAN or otherwise)?
>
> Many thanks
> Regards,
> Hugo
as others have pointed out, there are no I/O statistics in top. But you
might want to try sysstat to get an impression of what is going on in
your system.
Get it here: http://www.maier-komor.de/sysstat.html
Tom
|
|
0
|
|
|
|
Reply
|
Thomas
|
11/16/2006 7:35:52 PM
|
|
|
4 Replies
737 Views
(page loaded in 0.1 seconds)
Similiar Articles: top, prstat and prstat -m - comp.unix.solaris... comp.hardware top, prstat and prstat -m - comp.unix.solaris Use top, mpstat, iowait command to ... ... memory usage and free memory space, etc. 2) Top ... up the most CPU, disk ... Average wait time per IO request and ... - comp.unix.solaris ...Wait command - comp.unix.solaris 'Top' command and disk io readings - comp.unix.solaris ... ... PID USERNAME LWP PRI NICE SIZE RES STATE TIME CPU COMMAND ... disk io benchmark - comp.unix.solarisdisk io benchmark - comp.unix ... reset': retrying command ... > this is your disk @0,1 on io board ... Much of that is media read time (disk or network time ... Otherwise, IO ... Disk benchmark (sol10) - comp.unix.solaris... Any idea how I would get reproducable results for read ... disk io benchmark - comp.unix.solaris Disk benchmark ... Finding memory leaks through ps command - comp.unix ... iostat -En does not show serial number - comp.unix.solaris ...TOP iowait vs iostat; 100% reading accurate? - comp.unix.solaris ... iostat ... What means LWP column in the "top" command? Number ... throughput test; IOstat & finding the IO ... Identifying processes that are heavy IO users. - comp.unix.admin ...... on a script to identify IO usage on a high IO ... top: 50% iowait, 900M swap in use - comp.sys.sun ... Directly Reading/Writing to Disk - comp.lang.asm.x86 Can a kernel mode ... ioscan on HP Unix 11.11 - comp.sys.hp.hpux... We do not change fstab) and Boot IO ... INIT failed Error > Status -7 Tape reader has a conflict with the existing disk. ... HP-UX "ln -s" command - comp.sys.hp.hpux Posix ... How to analysis high IOWait problem - comp.unix.solaris... wrote a doc on this a bit over a year ago, have a read ... Have you tried "iostat/sar" to see on which disk huge ... What means LWP column in the "top" command? Number of threads ... %b (percent busy) with iostat - comp.unix.solaris... range between 10 and 30 msecs (this is a massive read ... It might also be a disk IO issue, in which case iostat is ... why? - comp.unix.solaris Give us some output from top ... High await in iostat? - comp.os.linux.miscI thought that meant how fast the disk could run the command and ... to the home position, resulting in top-like output for disk ... Average wait time per IO request and ... Inspecting disk IO performance with fio | Linux.comfio was created to allow benchmarking specific disk IO workloads. It can issue its IO ... Run status group 0 (all jobs): READ: io=128MiB, aggrb=2,352KiB/s, minb=2,352KiB/s ... How do I find out Linux Disk utilization?Linux Performance Monitoring & Tuning hard disk with iostat command. ... I have had massive IO wait issues at times and ... How To: Double Linux disk read performance with ... 7/22/2012 4:13:22 PM
|