Informix 7.30 slows down

This is a multi-part message in MIME format. ------=_NextPart_000_0010_01C001ED.8DF51520 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Has anyone heard of the following problem. Tech Support have not been able to locate a problem or give a reason. Informix 7.30.UC2 Workgroup Edition SCO OpenServer 5.0.4 Our client was running with Informix AIO, NUMAIOVPS=16. We reconfigured their system and turned KAIO on. When we reduced NUMAIOVPS to 2, the system ran normally for about 3 - 4 hours and then started going noticably slow. Observing their system, everything looked to be going in slow motion, with no resource contention anywhere - including normal checkpoints of 1 - 2 seconds. As a measure of the performance degradation, they were running a job that was processing 11 products per minute. We stopped and restarted Informix, restarted this job and it then ran 170 products per minute. Their system started going slow after a few hours. We eventually located the problem after changing NUMAIOVPS back to 16 and having no further issues. This is contrary to tuning NUMAIOVPS with KAIO on. TIA ------=_NextPart_000_0010_01C001ED.8DF51520 Content-Type: text/plain; name="onconfig.txt" Content-Transfer-Encoding: quoted-printable Content-Disposition: attachment; filename="onconfig.txt" =0A= Informix Dynamic Server Version 7.30.UC2 -- On-Line -- Up 1 days = 18:06:20 -- 425984 Kbytes=0A= =0A= Configuration File: /home/informix/etc/onconfig.ows=0A= #************************************************************************= **=0A= #=0A= # INFORMIX SOFTWARE, INC.=0A= #=0A= # Title: onconfig.ows (/home/informix/etc/onconfig.ows)=0A= # Description: INFORMIX-OnLine Configuration Parameters=0A= #=0A= # 15/12/99 D.A - SHMBASE 0X10000000L 0X82000000L=0A= # MULTIPROCESSOR 0 1=0A= # NUMCPUVPS 1 2 =0A= # NETTYPE tlitcp,1,80,CPU tlitcp,2,80,CPU=0A= # DBSERVERALIASES and associated NETTYPE are commented out=0A= # BUFFERS 70000 65000=0A= # SHMVIRTSIZE 50000 58000=0A= # NETTYPE reverted - Engine failed to start!!!=0A= # 14/01/00 M.W SHMVIRTSIZE 58000 70000=0A= # BUFFERS 65000 120000=0A= # LRU_MAX_DIRTY 5 3=0A= # NUMAIOVPS 4 10=0A= # CLEANERS 8 10=0A= # LRU 8 10=0A= # 15/02/00 M.W SHMVIRTSIZE 70000 78000=0A= # LRU_MAX_DIRTY 3 2=0A= # CLEANERS 10 20=0A= # LRU 10 20=0A= # BUFFERS 120000 160000=0A= # 20/05/00 M.W New Disk Config=0A= # PHYSFILE 60000 80000=0A= # NUMAIOVPS 10 15=0A= # CLEANERS 20 40=0A= # LRU 20 40=0A= # NOAGE 0 1=0A= # RESIDENT 0 1=0A= # RA_THRESHOLD 0 16=0A= # RA_PAGES 0 24=0A= # DBSPACETEMP rootdbs tempdb1,2,3=0A= # 21/05/00 M.W NOAGE 1 0 (machine hung)=0A= # RESIDENT 1 0=0A= # NUMAIOVPS 16 2 KAIO ON=0A= # 12/06/00 M.W NUMAIOVPS 2 16=0A= # 23/06/00 M.W NUMAIOVPS 16 2=0A= # 01/07/00 M.W NUMAIOVPS 2 16=0A= #************************************************************************= **=0A= # Root Dbspace Configuration=0A= ROOTNAME rootdbs # Root dbspace name=0A= ROOTPATH /dev/rootdbs =0A= ROOTOFFSET 0 # Offset of root dbspace into device = (Kbytes)=0A= ROOTSIZE 512000 # Size of root dbspace (Kbytes)=0A=# Disk Mirroring Configuration Parameters=0A= MIRROR 0 # Mirroring flag (Yes =3D 1, No =3D 0)=0A= MIRRORPATH =0A= MIRROROFFSET 0 # Offset into mirrored device (Kbytes)=0A=# Physical Log Configuration=0A= PHYSDBS rootdbs # Location (dbspace) of physical log=0A= PHYSFILE 80000 # Physical log file size (Kbytes)=0A=# Logical Log Configuration=0A= LOGFILES 154 # Number of logical log files=0A= LOGSIZE 2000 # Logical log size (Kbytes)=0A= # Diagnostics=0A=MSGPATH /home/informix/edl.log # System message log file path=0A= CONSOLE /dev/console # System console message path=0A= ALARMPROGRAM /home/informix/etc/log_full.sh # Alarm program path=0A=# System Archive Tape Device=0A= TAPEDEV /dev/rct0 =0A= #TAPEDEV /dev/null=0A= TAPEBLK 128 # Tape block size (Kbytes)=0A= TAPESIZE 15000000 # Maximum amount of data to put on tape = (Kbytes)=0A= # Log Archive Tape Device=0A= LTAPEDEV /dev/rct1 =0A= #LTAPEDEV /dev/null=0A= LTAPEBLK 16 # Log tape block size (Kbytes)=0A= LTAPESIZE 4000000 # Max amount of data to put on log tape = (Kbytes)=0A= # Optical=0A=STAGEBLOB # INFORMIX-OnLine/Optical staging area=0A= # System Configuration=0A= SERVERNUM 10 # Unique id corresponding to a OnLine = instance=0A= DBSERVERNAME edl # Name of default database server=0A= DBSERVERALIASES edl_shm # List of alternate dbservernames=0A= NETTYPE tlitcp,2,50,CPU =0A= NETTYPE ipcshm,1,20,NET =0A= DEADLOCK_TIMEOUT 60 # Max time to wait of lock in = distributed env.=0A= RESIDENT 0 # Forced residency flag (Yes =3D 1, No = =3D 0)=0A= MULTIPROCESSOR 1 # 0 for single-processor, 1 for = multi-processor=0A= NUMCPUVPS 2 # Number of user (cpu) vps=0A= SINGLE_CPU_VP 0 # If non-zero, limit number of cpu vps = to one=0A= NOAGE 0 # Process aging=0A= AFF_SPROC 0 # Affinity start processor=0A= AFF_NPROCS 0 # Affinity number of processors=0A= # Shared Memory Parameters=0A= LOCKS 100000 # Maximum number of locks=0A= BUFFERS 160000 # Maximum number of shared buffers=0A= NUMAIOVPS 16 # Number of IO vps=0A= PHYSBUFF 64 # Physical log buffer size (Kbytes)=0A= LOGBUFF 32 # Logical log buffer size (Kbytes)=0A=LOGSMAX 200 # Maximum number of logical log files=0A= CLEANERS 40 # Number of buffer cleaner processes=0A= SHMBASE 0x82000000 # Shared memory base address=0A= SHMVIRTSIZE 81920 # initial virtual shared memory segment = size=0A= SHMADD 8192 # Size of new shared memory segments = (Kbytes)=0A= SHMTOTAL 0 # Total shared memory (Kbytes). = 0=3D>unlimited=0A= CKPTINTVL 900 # Check point interval (in sec)=0A= LRUS 40 # Number of LRU queues=0A= LRU_MAX_DIRTY 2 # LRU percent dirty begin cleaning limit=0A= LRU_MIN_DIRTY 1 # LRU percent dirty end cleaning limit=0A= LTXHWM 40 # Long transaction high water mark = percentage=0A= LTXEHWM 50 # Long transaction high water mark = (exclusive)=0A= TXTIMEOUT 0x12c # Transaction timeout (in sec)=0A= STACKSIZE 32 # Stack size (Kbytes)=0A= # System Page Size=0A= # BUFFSIZE - OnL