Error -25573 OS error 76

Hello all: I have been having a slow down in connection times to our database over the past two days. The problem is intermittent, it has happened the past two mornings. The affect is users trying to make connections to the database are taking 1 to 3 minutes to make the initial connection to the database. Once the connection is made all other functions work properly. The following error has been logged to the online log. There has been 13 occurrences 1/19 and 129 occurrences 1/20. 09:25:40 listener-thread: err = -25573: oserr = 76: errstr = :Networkdriver cannot accept a connection on the port. System error = 76. The system error 76 says, "The socket is not in the connected state" I have opened a case with Informix Technical support # 928201 and IBM support #54305 Branch 004.. Informix has dialed into my server and run their trace utility to capture the error as it was occurring. The analysis of the trace indicates the following chain of events: Request comes into the socket for service. Informix listen thread sends the ack system call to accept the request. The ack system call returns -1. It is unclear (To me) exactly what the chain of events occur after that failure. IBM support says this is a normal condition and AIX is only reporting the fact that the connection is no longer there. IBM has logged two incidents of the error with two applications. The following is the resolutions of these two cases: Ondemand was the app for the client: 76 = client connection has abandoned an existing socket connection. An rst tcp packet has been transferred from the pc to the RISC server (that data can be capture in an IP trace or tcp dump). The app can disregard the error 76 and release the socket and put it back in the listen state. They talked to the Ondemand developers to render it as a nonfatal error. Oracle listener was at 7.3.3.5 needed to be at 7.3.4 or Oracle 8.0.3 or higher The IBM technician characterized the conversation like this. Someone calls you on the telephone. You are busy so the phone continues to ring. You finely make it to the phone and as soon as you pick up the receiver the other party hangs up. You say hello and continue to say hello, hello until you give up and hang up the receiver so the next call can be answered. During the time you are saying hello, hello new calls are being queued up on you call waiting. IBM's feeling is that the Informix listener thread should handle the 76 error by resetting the socket to the listen state. My gut feeling is that there is a client on out network that may have a bad NIC or something that is causing the problem to occur. I also believe Informix should handle this drop gracefully and not cause the delays. Any insight on this problem would be greatly appreciated. Joe Miller Administaff Inc. Joe_Miller@Administaff.com My environment is : Server: RS/6000 S7A, 12 Processors, 6GB Ram, AIX 4.3.2 Clients: Windows NT, Powerbuilder, 4gl on AIX. Onconfig: ROOTNAME rootdbs # Root dbspace name ROOTPATH /dev/rrdbs # Path for device containing rootdbspace ROOTOFFSET 0 # Offset of root dbspace into device (Kbytes) ROOTSIZE 307200 # Size of root dbspace (Kbytes) # Disk Mirroring Configuration Parameters MIRROR 0 # Mirroring flag (Yes = 1, No = 0) MIRRORPATH # Path for device containing mirroredroot MIRROROFFSET 0 # Offset into mirrored device (Kbytes) # Physical Log Configuration PHYSDBS rootdbs # Location (dbspace) of physical log PHYSFILE 102400 # Physical log file size (Kbytes) # Logical Log Configuration LOGFILES 282 # Number of logical log files LOGSIZE 2048 # Logical log size (Kbytes) # Diagnostics MSGPATH /usr/ifmx/online.log # System message log file path CONSOLE /usr/ifmx/cons.log # System console message path #ALARMPROGRAM /usr/ifmx/etc/log_full.sh # Alarm program path SYSALARMPROGRAM /usr/ifmx/etc/evidence.sh # System Alarm program path TBLSPACE_STATS 1 # System Archive Tape Device TAPEDEV /dev/rmt0 # Tape device path TAPEBLK 128 # Tape block size (Kbytes) TAPESIZE 73400320 # Maximum amount of data to put on tape (Kbytes) # Log Archive Tape Device LTAPEDEV /dev/rmt0 # Log tape device path LTAPEBLK 128 # Log tape block size (Kbytes) LTAPESIZE 73400320 # Max amount of data to put on log tape (Kbytes) # Optical STAGEBLOB # Informix Dynamic Server/Optical staging area # System Configuration SERVERNUM 1 # Unique id corresponding to a DynamicServer instance DBSERVERNAME kingdbs1_l # Name of default database serverDBSERVERALIASES kingdbs1,bigboy,bigboy_l # List of alternate dbservernames NETTYPE ipcshm,1,100,NET # Configure poll thread(s) fornettype NETTYPE soctcp,7,200,CPU # Configure poll thread(s) fornettype DEADLOCK_TIMEOUT 60 # Max time to wait of lock indistributed env. RESIDENT 0 # Forced residency flag (Yes = 1, No = 0) MULTIPROCESSOR 1 # 0 for single-processor, 1 formulti-processor NUMCPUVPS 11 # Number of user (cpu) vps SINGLE_CPU_VP 0 # If non-zero, limit number of cpu vpsto one NOAGE 1 # Process aging AFF_SPROC 0 # Affinity start processor AFF_NPROCS 0 # Affinity number of processors # Shared Memory Parameters LOCKS 500000 # Maximum number of locks BUFFERS 175000 # Maximum number of shared buffers NUMAIOVPS 1 # Number of IO vps PHYSBUFF 64 # Physical log buffer size (Kbytes) LOGBUFF 64 # Logical log buffer size (Kbytes)LOGSMAX 600 # Maximum number of logical log files CLEANERS 20 # Number of buffer cleaner processes SHMBASE 0x30000000 # Shared memory base address SHMVIRTSIZE 1000000 # initial virtual shared memorysegment size SHMADD 64768 # Size of new shared memory segments (Kbytes) SHMTOTAL 0 # Total shared memory (Kbytes). 0=>unlimited CKPTINTVL 900 # Check point interval (in sec) LRUS 24 # Number of LRU queues LRU_MAX_DIRTY 2 # LRU percent dirty begin cleaninglimit LRU_MIN_DIRTY 1 # LRU percent dirty end cleaning limit LTXHWM 50 # Long transaction high water markpercentage LTXEHWM 60 # Long transaction high water mark (exclusive) TXTIMEOUT 0x12c # Transaction timeout (in sec) STACKSIZE 32 # Stack size (Kbytes) # System Page Size # BUFFSIZE - Dynamic Server no longer supports this configuration paramet