How the Core Dump Subsystem works
The following illustration shows the key relationships between different
components used to create the core dump in the case of an application failure
or a system crash:
The core dump server observes the threads/processes in the observed set.
It collects failure information for the failed threads or processes to create
the core dump. When an application failure event is reported, the core dump
server requests for failure information from the Debug Security Server which is
part of run-mode debug subsystem.
In the case of a system failure, the System Crash Monitor (SCM) receives
notification about the occurrence of the event. SCM collects the core dump and
passes it to the reserved crash partition. The type of information that SCM
collects depends on the configuration data that it receives from the CDS. This
configuration data can be set through the core dump UI. In addition, you can
use the UI to process the information and generate a core dump file. For
details on setting the configuration using the TechView UI, see
Using the TechView UI to Capture Application Failure Information and
Using the TechView UI to Capture System Crash Information.
You can also develop a UI application to configure CDS. For details, see
Developing a Configuration UI for the Core Dump Subsystem.