StarlingX Fault Management Project¶

The starlingx/fault project provides fault management.

Overview¶

Fault management is a component of StarlingX project that detects, classifies, and notifies different behaviors occurring in the infrastructure.

Undesirable behaviors in the system are detected as alarms that usually need corrective actions intermediately. The severity is determined according to a classification (e.g. critical, warning, and others) that provide the operator with more information about the system’s condition. The classification also provides which system component is affected (e.g. network, storage, security, and so forth).

Additionally, fault management (FM) handles expected behaviors organized in events. These behaviors are classified and show relevant information for the administrator that might require an action. Fault Management is a useful service to monitor the infrastructure’s health and allows you to action according to priorities.

Fault management is comprised of a group of sub-projects that work together to offer the FM service. Following is a brief description of FM’s role:

Sub-project	Description
fm-api	Application API used by software applications to raise, clear and audit the alarms and events.
fm-common	Includes libraries used by the fm-mgr. Additionally, fm-mgr includes a client that interacts with fm-mgr throughout a socket.
fm-mgr	Creates, reads, and deletes faults. fm-mgr can access the database.
fm-rest-api	Through use of a RESTful API, allows information on events and alarms from the database to be reported. For more information, see the API reference.
python-fmclient	Allows use of the Fault Management RESTful API.

StarlingX services can use Fault Management through a client that at the same time uses the API to handle the information in the database. The following diagram describes the process: