Considerations for communications-related system values

System values, such as system date and library list, control information for the operation of certain parts of the system. You can change the system value to define your working environment.

The following information explains more about each of the system values for communications error recovery.

• QCMNARB (communications arbiters): controls the number of communication arbiter system jobs that are available to process communications functions.
  • Do not set this value to zero unless directed by software service. If this system value is set to zero, the work is performed in the QSYSARB and QLUS system jobs as opposed to being performed by the communication arbiters.
  • The QCMNARB system value supports the following values: *CALC, 0-99.
  • *CALC is the default setting for this system value. The system determines the number of jobs based on the system's HW configuration.
  • Consider having more than one QCMNARB job if there is an excessive amount of these system activities.
  • A change to this value requires an initial program load (IPL) of the system for it to take affect.
• QPASTHRSVR (pass-through servers): controls how many pass-through server jobs are available for processing display station pass-through requests.
  • The default setting of this system value is calculated based on the hardware configuration of your system.
  • Consider multiple pass-through server jobs in error recovery situations to make the system quicker.
  Note:
  Setting the QPASTHRSVR value to 0 is not recommended. The QPASTHRSVR value of 0 is intended for use in migrating from the use of communications jobs for the 5250 target display station pass-through function to use the pass-through server jobs.
• QCMNRCYLMT (communications recovery limit): controls the number of automatic recovery attempts to make. It also controls when an inquiry message is sent to the system operator when the specified number of recovery attempts has been reached.
  • If the CMNRCYLMT parameter value is specified as *SYSVAL for a network interface description, a line description, or a controller description, then the QCMNRCYLMT value is also used. These parameter values also contain a count limit and time interval.

    The count limit can be 0 (no recovery attempted) to 99. The time interval can be 0, or a value from 1 to 120 (minutes). A count limit of 0, and a time interval of more than 0 effectively disables automatic second-level error recovery. This may cause the devices and controllers to go into recovery pending (RCYPND) state, and require operator intervention. A count limit of more than 0, and a time interval of 0 allows automatic second-level error recovery continuously. However, this is not recommended.

    Note:
    To avoid looping recovery, keep the number of retries small; you do not want time to expire before the number of retries are exceeded. Otherwise, you will end up in infinite recovery.
• QDEVRCYACN (device I/O recovery action): Controls the recovery action to take for the job when a device error is encountered on a reading and writing operation on the *REQUESTER device for interactive jobs.

• For more information about iSeries™ communications, see the Communications Management and Work Management books.
• For more Information about pass-through server jobs, see the Remote Work Station Support book.