RE: KAIO and HPUX 11.0 with multiple Informix 7.31 instances

Don't know if it's true for ver. 7.31 also, but for 9.21 the resident memory segment MUST be locked in (i.e., resident) for KAIO to work. I agree with David -- with that many instances, there's a good chance that the resident segments for at least *some* of them are not getting locked in. Check the output of 'onstat -g seg' for each instance to see it they are resident. Just because you "requested" that they be resident in your onconfig does not necessarily mean that the O/S was *able* to make them resident. Also, if you're getting additional virtual segments (i.e., more than one) on some of the instances -- well that almost *always* causes headaches for HP-UX. :-/ HTH, Paul Mosser -----Original Message----- From: David Williams [mailto:djw@smooth1.demon.co.uk] Sent: Tuesday, January 30, 2001 4:53 PM To: informix-list@iiug.org Subject: Re: KAIO and HPUX 11.0 with multiple Informix 7.31 instances In article <20010129184239.11105.00001252@ng-cg1.aol.com>, Pxllr <pxllr@aol.com> writes >Has anyone seen a problem with multiple informix engines on HPUX 11.0? My >question is mainly that it appears our main production engine is set to use >KAIO and it looks to me like it is configured to use the entire HPUX box >(K260-4way with 2.5GB RAM). The onconf.dbname for each instance is set the >exact same. This might be great for one instance, but whenever I try to bring >up a second instance, I start getting bus error, core dump. HP is stumped, and >I think it is a configuration problem in the engines. >Also swapinfo is going from 11% to 89%. After a reboot, 11% then it goes up >steadily till about 89% by Sunday, and I reboot Monday morning and start the >process again. This may or may not be connected. >Has anyone exerienced this, and what fixes it? > ># System Configuration > >SERVERNUM 2 # Unique id corresponding to a DynamicServer >instance You can't have two instances with the same servernum! >DBSERVERNAME cras # Name of default database server >DBSERVERALIASES crasitcp # List of alternate dbservernames >NETTYPE ipcshm,1,250,CPU # Configure poll thread(s) for nettype >NETTYPE soctcp,1,50,NET # Configure poll thread(s) for nettype >DEADLOCK_TIMEOUT 60 # Max time to wait of lock in distributed env. >RESIDENT 1 # Forced residency flag (Yes = 1, No = 0)> RESIDENT=1 means Informix tries to pin the shared memory into physical memory i.e. make the shared memory unswapable. You probably do not have enough physical memory to maek this possible. Try boot the seconds instance with oninit -v this undocumented option shows more info on how online starts up. >MULTIPROCESSOR 1 # 0 for single-processor, 1 formulti-processor >NUMCPUVPS 3 # Number of user (cpu) vps >SINGLE_CPU_VP 0 # If non-zero, limit number of cpu vps toone > >NOAGE 1 # Process aging >AFF_SPROC 0 # Affinity start processor >AFF_NPROCS 0 # Affinity number of processors> ># Shared Memory Parameters > >LOCKS 256000 # Maximum number of locks >BUFFERS 10000 # Maximum number of shared buffers >NUMAIOVPS 14 # Number of IO vps >PHYSBUFF 32 # Physical log buffer size (Kbytes) >LOGBUFF 32 # Logical log buffer size (Kbytes)>LOGSMAX 100 # Maximum number of logical log files >CLEANERS 4 # Number of buffer cleaner processes >SHMBASE 0x0 # Shared memory base address >SHMVIRTSIZE 92000 # initial virtual shared memory segmentsize >SHMADD 8192 # Size of new shared memory segments (Kbytes) >SHMTOTAL 0 # Total shared memory (Kbytes). 0=>unlimited >CKPTINTVL 900 # Check point interval (in sec) >LRUS 10 # Number of LRU queues >LRU_MAX_DIRTY 10 # LRU percent dirty begin cleaning limit >LRU_MIN_DIRTY 5 # LRU percent dirty end cleaning limit >LTXHWM 45 # Long transaction high water markpercentage >LTXEHWM 55 # Long transaction high water mark (exclusive) >TXTIMEOUT 0x12c # Transaction timeout (in sec) >STACKSIZE 32 # Stack size (Kbytes)> >TIA >Larry Romo >Programmer and other duties as assigned, such as HPUX systems Admin and >Informix DBA. -- David Williams