Re: System is slow ...
HW: Intel Pentium, 350MHZ, 128MB RAM, 5GB HD.
Results from different onstat command ...
$ onstat -p
INFORMIX-OnLine Version 7.20.UC2 -- On-Line -- Up
20:18:11 -- 8936 Kbytes
Profile
dskreads pagreads bufreads %cached dskwrits pagwrits
bufwrits %cached
244228 244246 1392999 82.47 333 595
18277 98.18
isamtot open start read write rewrite
delete commit rollbk
369980 1745 2900 350564 8409 12
0 108 0
ovlock ovuserthread ovbuff usercpu syscpu
numckpts flushes
0 0 0 22.20 12.30 9
480
bufwaits lokwaits lockreqs deadlks dltouts ckpwaits
compress seqscans
16280 0 659431 0 0 1
136 171
ixda-RA idx-RA da-RA RA-pgsused lchwaits
45876 191 175567 221497 8
------------------------------------------------
$ onstat -g iov
INFORMIX-OnLine Version 7.20.UC2 -- On-Line -- Up
20:18:29 -- 8936 Kbytes
AIO I/O vps:
class/vp s io/s totalops dskread dskwrite dskcopy
wakeups io/wup
msc 0 i 0.0 75 0 0 0
63 1.2
aio 0 i 0.2 12691 12621 36 0
6822 1.9
aio 1 i 0.2 12147 12090 50 0
7050 1.7
aio 2 i 0.1 10817 10787 30 0
9648 1.1
aio 3 i 0.1 10243 10217 26 0
9812 1.0
aio 4 i 0.1 9913 9892 21 0
9765 1.0
aio 5 i 0.1 9730 9710 20 0
9687 1.0
aio 6 i 0.1 9289 9267 22 0
9285 1.0
aio 7 i 0.1 9016 8994 22 0
8995 1.0
aio 8 i 0.1 9652 9627 25 0
8548 1.1
aio 9 i 0.1 9491 9469 22 0
8177 1.2
pio 0 i 0.0 21 0 21 0
21 1.0
lio 0 i 0.0 13 0 13 0
14 0.9
-----------------------------------------
$ onstat -R
INFORMIX-OnLine Version 7.20.UC2 -- On-Line -- Up
20:18:50 -- 8936 Kbytes
8 buffer LRU queue pairs
# f/m length % of pair total
0 f 24 96.0% 25
1 m 1 4.0%
2 f 24 96.0% 25
3 m 1 4.0%
4 f 22 88.0% 25
5 m 3 12.0%
6 f 23 92.0% 25
7 m 2 8.0%
8 F 21 84.0% 25
9 m 4 16.0%
10 f 24 96.0% 25
11 m 1 4.0%
12 f 23 92.0% 25
13 m 2 8.0%
14 f 22 88.0% 25
15 m 3 12.0%
17 dirty, 200 queued, 200 total, 256 hash buckets,
2048 buffer size
start clean at 60% (of pair total) dirty, or 15 buffs
dirty, stop at 50%
----------------------------------------------
$ onstat -F
INFORMIX-OnLine Version 7.20.UC2 -- On-Line -- Up
20:19:07 -- 8936 Kbytes
Fg Writes LRU Writes Chunk Writes
0 38 242
address flusher state data
10438448 0 I 0 = 0X0
states: Exit Idle Chunk Lru
--------------------------------------------------
$ onstat -b | tail -5
Buffers
address userthread flgs pagenum memaddr nslots
pgflgs xflgs owner waitlist
17 modified, 200 total, 256 hash buckets, 2048 buffer
size
---------------------------------------------------
$ onstat -l
INFORMIX-OnLine Version 7.20.UC2 -- On-Line -- Up
20:19:39 -- 8936 Kbytes
Physical Logging
Buffer bufused bufsize numpages numwrits pages/io
P-1 0 16 278 21 13.24%
phybegin physize phypos phyused %used
10003f 1000 122 16 1.60
Logical Logging
Buffer bufused bufsize numrecs numpages numwrits
recs/pages pages/io
L-2 1 16 580 18 13
32.2 1.4
address number flags uniqid begin size
used %used
100470fc 1 U---C-L 19 100427 500
379 75.80%
10047118 2 U-B---- 14 10061b 500
500 100.00%
10047134 3 U-B---- 15 10080f 500
500 100.00%
10047150 4 U-B---- 16 100a03 500
500 100.00%
1004716c 5 U-B---- 17 100bf7 500
500 100.00%
10047188 6 U-B---- 18 100deb 500
500 100.00%
Waiting for your valueable comments ...
Thanks in advance.
--- Obnoxio The Clown <obnoxio@hotmail.com> wrote:
> From: Shahid Mehmood <theixdba@yahoo.com>
> >
> >OS: SCO UNIX -- OPEN SERVER 5.0
> >IX: OWS 7.20.UC2
> >
> >I have one box with OWS installed on it. This
> instance
> >contains 4 dbspaces, and 4 respective databases.
> All
> >the users connect to their respective databases
> >through application written using Informix 4GL
> (c-code
> >version). The users are complaining that the system
> >response is slow.
> >
> >Any suggestions, as how to improve the performance?
> I
> >am waiting anxiously for your valueable replies,
> and
> >comments.
> >
> >Thanks.
> >
> >My onconfig.ows file is as follows:
> >
> >$ cat onconfig.ows>
>#**************************************************************************
> >#
> ># INFORMIX SOFTWARE, INC.
> >#
> ># Title: onconfig.ows
> ># Description: INFORMIX-OnLine Configuration
> >Parameters
> >#
>
>#**************************************************************************
> >
> ># Root Dbspace Configuration
> >
> >ROOTNAME rootdbs # Root dbspace name
> >ROOTPATH /u/IFMXDATA/AKU/ROOTDBS.000
> >ROOTOFFSET 0 # Offset of root> >dbspace into device (Kbytes)
> >ROOTSIZE 20000 # Size of root> dbspace
> >(Kbytes)
> >
> ># Disk Mirroring Configuration Parameters
> >
> >MIRROR 0 # Mirroring flag (Yes =
> 1,
> >No = 0)
> >MIRRORPATH
> >MIRROROFFSET 0 # Offset into> mirrored
> >device (Kbytes)
> >
> ># Physical Log Configuration
> >
> >PHYSDBS rootdbs # Location
> (dbspace)
> >of physical log
> >PHYSFILE 2000 # Physical log file
> >size (Kbytes)>
> Move your physical log to it's own dbspace, on it's
> own spindle.
>
> ># Logical Log Configuration
> >
> >LOGFILES 6 # Number of logical> >log files
> >LOGSIZE 1000 # Logical log size
> >(Kbytes)> >
> ># Diagnostics
> >
> >MSGPATH /u/usr/informix/etc/online.log #
> >System message log file path
> >CONSOLE /dev/console # System console> >message path
> >ALARMPROGRAM /u/usr/informix/etc/log_full.sh
> #
> >Alarm program path
> >
> ># System Archive Tape Device
> >
> >TAPEDEV /dev/rStp0
> >TAPEBLK 16 # Tape block size
> >(Kbytes
Only have 200 buffers is pretty tiny (i.e. 400KB on a 128MB system), but you only have 1 disk as well, so unless the database is so small as to fit in memory, you are going to be thrashing the disk between writing out pages and logs etc. Note in general it isn't helpful, to have more cleaners than disks so 8 (as I think was previously suggested) is liley overkill, 1 or 2 would be plenty. Particularly with a 1 cpu 1 disk system. -jrg Sent via Deja.com http://www.deja.com/ Before you buy.