For

CIMPLICITY Interprocess Communication (IPC) System

Purpose

To modify the parameters IPC uses to determine what is "reasonable" growth in the use of system memory.

CIMPLICITY Interprocess Communication (IPC) will accommodate "reasonable" periods of temporary growth in the use of system memory, yet try to keep errant clients from causing IPC Memory growth to consume all resources..

There are two sets of global parameters. The first set is for On-Node message "ports"; the other set is for Off-Node communications through "sockets".

On-Node Message "Ports"

Global Parameter

Specifies:

PORT_PERIOD_SECS

The number of seconds in a calculation period.

The default is 5.

PORT_GROWTH_PERIODS

The number of periods to allow that message production exceeds message consumption. After this number of periods messages will be dropped. Defaults to 6.

PORT_NONE_CONSUMED

The number of periods to allow with no message consumption. Defaults to 3.

When the queue of port overflows and exceeds the above parameters the following message will be logged:

Router to Application queue overflow on port

If you see the above message tuning the above parameters may help. Typically the problem occurs because data is being produced faster than it can be consumed and changing these parameters will only delay the problem not solve it. Typically the solution to the problem is to slow down production (e.g. reduce scan rates) or speed up production (e.g. faster computer, more RAM)

Off node Communications through "Sockets"

Global Parameter

Specifies

SOCK_PERIOD_SECS

The number of seconds in a calculation period. Defaults to 15.

SOCK_GROWTH_PERIODS

The number of periods to allow that message production exceeds message consumption. After this number of periods messages will be dropped. Defaults to 6.

SOCK_NONE_CONSUMED

The number of periods to allow with no message consumption. Defaults to 3.

When the queue of the "socket" overflows and exceeds the above parameters the following message will be logged:

Router dropped off node message to…

If you see the above message tuning the above parameters may help. Typically the problem occurs because data is being produced faster than it can be consumed and changing these parameters will only delay the problem not solve it. Typically the solution to the problem is to slow down production (e.g. reduce scan rates) or speed up production (e.g. faster network, faster computers, more RAM).

Global parameters.