EMS Manual

Event Routing
EMS Manual426909-005
16-4
Distributor Generated Messages
If a DEFINE is not specified, all distributor generated events except
ZEMS^EVT^BURST^START and ZEMS^EVT^BURST^END are written to $0.
The burst start and burst end messages are written to the OUT file of the distributor, as
specified in the EMSDIST run option, OUT. If the EMSDIST run option OUT is specified
but does not match the hometerm, burst start and burst end messages are written to
the OUT file as a text error list. If the EMSDIST run option OUT is specified without a
value, burst start and burst end messages are written to the hometerm of the
distributor as a text error list. If the value of the EMSDIST run options OUT and TERM
are the same, and neither the =_ems_system_log or =_ems_alternate_log defines
exist, burst start and burst end messages are not written.
Distributor Generated Messages
Previously, if an OUT file was not explicitly specified, all printing or forwarding
distributor generated error or warning messages were sent to the home terminal. This
included messages that were generated after successful distributor startup. These
messages were also converted to equivalent events and sent to $0. Because
messages sent to the OUT file are not identified as to what distributor they are
associated with, they might have little meaning for an operator. This is especially true
for the new destination process error messages.
The current strategy is to send only messages that report distributor startup problems
to the home terminal. If an OUT file other than the home terminal is given, all
messages are recorded as before. Also, if OUT is specified without a file name, all
messages are sent to the home terminal. Event generation is not affected.
The distributor generates its own event and sends it to up to two user selectable
destinations when:
An event cannot be written to the destination because of a file system error (first
occurrence).
The destination process can not be started (first occurrence).
The destination process was successfully started after previous failure and
subsequent retries.
All retries have failed and resulted in repeated timeouts over a period of 24 hours.