onstat -g iov
Posted in 2009
Topics: General Discussion
No doubt the real answer is RTFM. But, can anyone give me or point me to a
good description of the i/o -wakeup ratio, and why 0.5 is "good"?:
onstat -g iov -r 5
IBM Informix Dynamic Server Version 10.00.FC8W2 -- On-Line (Prim) -- Up 81
days 06:27:37 -- 46892464 Kbytes
AIO I/O vps:
class/vp s io/s totalops dskread dskwrite dskcopy wakeups io/wup
errors
kio 0 i 41.6 36343 21594 14749 0 57257 0.6 0
kio 1 i 37.9 33161 13908 19253 0 49383 0.7 0
kio 2 i 27.8 24334 14429 9905 0 38749 0.6 0
kio 3 s 33.7 29439 16651 12788 0 46029 0.6 0
kio 4 i 15.3 13339 7557 5782 0 20654 0.6 0
kio 5 i 19.5 17059 10091 6968 0 27466 0.6 0
kio 6 i 0.9 762 100 662 0 798 1.0 0
kio 7 i 2.0 1721 429 1292 0 2021 0.9 0
kio 8 i 6.9 6051 3385 2666 0 9600 0.6 0
kio 9 i 3.2 2788 1128 1660 0 3899 0.7 0
kio 10 i 1.0 833 63 770 0 930 0.9 0
kio 11 i 4.3 3768 2346 1422 0 5835 0.6 0
kio 12 i 11.0 9636 5736 3900 0 15246 0.6 0
kio 13 i 1.4 1222 230 992 0 1460 0.8 0
kio 14 i 1.4 1237 131 1106 0 1343 0.9 0
kio 15 i 1.1 929 194 735 0 1038 0.9 0
kio 16 i 1.4 1264 220 1044 0 1491 0.8 0
kio 17 i 2.7 2350 924 1426 0 3004 0.8 0
kio 18 i 1.0 869 131 738 0 953 0.9 0
kio 19 i 0.7 625 108 517 0 718 0.9 0
kio 20 i 0.7 596 166 430 0 696 0.9 0
kio 21 i 2.3 1990 98 1892 0 1774 1.1 0
kio 22 i 0.5 416 57 359 0 504 0.8 0
kio 23 i 0.6 514 58 456 0 650 0.8 0
kio 24 i 0.3 262 79 183 0 360 0.7 0
kio 25 i 0.6 500 71 429 0 591 0.8 0
kio 26 i 0.7 603 26 577 0 614 1.0 0
kio 27 i 0.2 206 64 142 0 303 0.7 0
msc 0 i 1.0 863 0 0 0 861 1.0 0
aio 0 i 0.3 294 98 0 0 294 1.0 0
aio 1 i 0.0 0 0 0 0 0 0.0 0
aio 2 i 0.0 0 0 0 0 0 0.0 0
pio 0 i 0.0 0 0 0 0 0 0.0 0
lio 0 i 0.0 0 0 0 0 0 0.0 0
thx
Neil
Neil Truby wrote:
> No doubt the real answer is RTFM. But, can anyone give me or point me
> to a good description of the i/o -wakeup ratio, and why 0.5 is "good"?:
>
> onstat -g iov -r 5
> IBM Informix Dynamic Server Version 10.00.FC8W2 -- On-Line (Prim) --
> Up 81 days 06:27:37 -- 46892464 Kbytes>
> AIO I/O vps:
> class/vp s io/s totalops dskread dskwrite dskcopy wakeups io/wup
> errors
> kio 0 i 41.6 36343 21594 14749 0 57257 0.6 0
> kio 1 i 37.9 33161 13908 19253 0 49383 0.7 0
> kio 2 i 27.8 24334 14429 9905 0 38749 0.6 0
> kio 3 s 33.7 29439 16651 12788 0 46029 0.6 0
> kio 4 i 15.3 13339 7557 5782 0 20654 0.6 0
> kio 5 i 19.5 17059 10091 6968 0 27466 0.6 0
> kio 6 i 0.9 762 100 662 0 798 1.0 0
> kio 7 i 2.0 1721 429 1292 0 2021 0.9 0
> kio 8 i 6.9 6051 3385 2666 0 9600 0.6 0
> kio 9 i 3.2 2788 1128 1660 0 3899 0.7 0
> kio 10 i 1.0 833 63 770 0 930 0.9 0
> kio 11 i 4.3 3768 2346 1422 0 5835 0.6 0
> kio 12 i 11.0 9636 5736 3900 0 15246 0.6 0
> kio 13 i 1.4 1222 230 992 0 1460 0.8 0
> kio 14 i 1.4 1237 131 1106 0 1343 0.9 0
> kio 15 i 1.1 929 194 735 0 1038 0.9 0
> kio 16 i 1.4 1264 220 1044 0 1491 0.8 0
> kio 17 i 2.7 2350 924 1426 0 3004 0.8 0
> kio 18 i 1.0 869 131 738 0 953 0.9 0
> kio 19 i 0.7 625 108 517 0 718 0.9 0
> kio 20 i 0.7 596 166 430 0 696 0.9 0
> kio 21 i 2.3 1990 98 1892 0 1774 1.1 0
> kio 22 i 0.5 416 57 359 0 504 0.8 0
> kio 23 i 0.6 514 58 456 0 650 0.8 0
> kio 24 i 0.3 262 79 183 0 360 0.7 0
> kio 25 i 0.6 500 71 429 0 591 0.8 0
> kio 26 i 0.7 603 26 577 0 614 1.0 0
> kio 27 i 0.2 206 64 142 0 303 0.7 0
> msc 0 i 1.0 863 0 0 0 861 1.0 0
> aio 0 i 0.3 294 98 0 0 294 1.0 0
> aio 1 i 0.0 0 0 0 0 0 0.0 0
> aio 2 i 0.0 0 0 0 0 0 0.0 0
> pio 0 i 0.0 0 0 0 0 0 0.0 0
> lio 0 i 0.0 0 0 0 0 0 0.0 0
>
> thx
> Neil
Well, near to "1" is good - i.e. wake up and do 1 thing.
More than 1 would mean that there is a chance that if more resources were available then that "extra bit" could have been done sooner.
Less than 1 would mean ... something else :-/ Something like "it was woken up, but not to do any work".
Anyways, that output seems to suggest you are running fine as regards I/O, note how your first 5 KIO VPs are taking *most* of the
I/O, the others are pretty dormant.
Neil Truby wrote: > No doubt the real answer is RTFM. But, can anyone give me or point me to a > good description of the i/o -wakeup ratio, and why 0.5 is "good"?: It's not "good" per se, it just means that you are doing less than 1 IO every time the IO thread gets fired up, which means that you have a sufficient number of IO threads configured. -- Cheers, Obnoxio The Clown http://obotheclown.blogspot.com -- This message has been scanned for viruses and dangerous content by OpenProtect(http://www.openprotect.com), and is believed to be clean.
Related threads
- Posting from the Informix-list
- Migrating from IDS 9.40.UC6 to 11.50.UC3
- Ip for a network session
- questions onstat -g