Skip to content

3.11 Diagnostics

lrrajesh edited this page Sep 14, 2016 · 1 revision

System level diagnostics will detect and report failures of critical resources, including memory, processors, network paths and I/O interfaces. The diagnostic routines will be capable of isolating hardware problems down to the Field Replaceable Unit (FRU) level in both the system and its peripheral equipment. The Sensys diagnostics framework supplies a consistent set of APIs for invoking the desired diagnostics functionality, while hiding the complexity of the implementation.

Normally, diagnostics can be launched by the system administrator via the octl tool. The octl diagnostics commands can invoke the desired diagnostic tests on a specific node or a list of nodes. The actual diagnostic tests will be executed on the compute node(s) quietly. Once the diagnostic is completed, compute nodes will send their result to the aggregator node, which logs the data into database. The diagnostic result logging feature is still under development.

Diagnostics Launch

Clone this wiki locally