Listed below are common errors that might occur when running a Plex application that connects through the Plex dispatcher on an IBM i machine. All of the errors below can be found in the QPRINT log file generated on the IBM i machine. This log file is located in either the OUTQ of the user profile that submits the dispatcher program, or in the OUTQ of the connecting user profile.
- Error on socket bind: errorno = 3420.
The error 3420 means the port is already in use.
Conditions that may cause this error include:
- the port specified was not available, in use or blocked by a firewall
- the TCP/IP or QINTER subsystem is not running
If this error is encountered, change the port number to use an available port or if the incoming port is blocked by a firewall, use the alternate YOBSYTCPDP dispatcher. (YOBSYTCP does not operate through a firewall.)
- Error on read of pinfo: errno = 3426
This error occurs when there is an abrupt disconnection between the PC and the IBM i.
Possible causes of the problem include:
- user has logged off the PC without first ending the Plex application
- user has powered off the PC without first ending Plex application
- network outage
- Error on connect to client socket: errno = 3447
Error 3447 means the host did not respond within the timeout period.
Possible causes of the problem:
- a period of inactivity between the PC and the IBM i caused connection to terminate
Verify the keepalive value in the IBM i CHGTCPA command is set to an appropriate value. The keepalive parameter is a value specifying the interval at which the IBM i communicates over the connection to keep the connection from timing out.