OpenHPI is an open-source software system providing an abstracted interface to managing computer hardware, typically for chassis and rack based servers.[1][2] It is production ready implementation of the Hardware Platform Interface specification from Service Availability Forum, complimenting existing hardware management standards. Founded in 2003, OpenHPI is maintained by the OpenHPI Project.[3][4]
OpenHPI provides resource modeling, sensor management, control, watchdog, inventory data associated with resources, abstracted system event log, hardware events/alarms, and a managed hot-swap interface.[5] It aims for Service Availability beyond High Availability (HA) expectations.[1][4]
The OpenHPI project was conceived by Carrier Grade Linux hardware experts,[3] and announced on the Linux kernel mailing list on 19 March 2003, by Andrea Brugger. OpenHPI was described as "a universal interface for creating resource system models, such as chassis and rack-based servers, but extendable for other domains such as clustering, virtualization, and simulation". It had modular hardware support implemented using a plugin architecture, the top-level OpenHPI implementation being independent of the underlying hardware.[6] Supporters include IBM, Intel, Samsung, HPE, and others technical equipment manufacturers.
The following features are supported by OpenHPI software:[2]
OpenHPI also provides a set of client programs as examples for typical HPI usage, for testing, or invocation from scripts. The hpi_shell is a command shell for calling HPI functions interactively.
The following table summarizes the main OpenHPI releases: