Synchronized debug information generation
In an approach for determining a location of failure between interconnects/controller, a computer collects debug information simultaneously at a plurality of nodes coupled to an interconnect. Subsequent to collecting debug information, the computer analyzes the debug information collected simultaneously thereby determining which end of the interconnect caused the failure.
1. A method for determining a location of failure between interconnects/controller, the method comprising:
determining, by a first system controller coupled to a second system controller and a plurality of node controllers of a server computer, an error in the server computer, wherein the server computer includes a plurality of service processors configured in a hierarchical architecture;
responsive to determining the error in the server computer, determining, by the first system controller, that a simultaneous dump should be performed by a plurality of both the second system controller and the plurality of node controllers;
responsive to determining that the simultaneous dump should be performed, broadcasting, by the first system controller, an alert to the plurality of both the second system controller and the plurality of node controllers to perform a simultaneous dump; and
receiving, by the first system controller, the simultaneous dump from the plurality of both the second system controller and the plurality of node controllers.
2. The method of claim 1 , wherein broadcasting, by the first system controller, an alert to the plurality of both the second system controller and the plurality of node controllers comprises signaling, by the first system controller, a programmable interrupt generator in the server computer to communicate with the plurality of system controllers and the plurality of node controllers.
3. The method of claim 1 , wherein broadcasting, by the first system controller, an alert to the plurality of both the second system controller and the plurality of node controllers comprises the first system controller broadcasting the error on an ethernet transport on which the plurality of both the second system controller and the plurality of node controllers reside.
4. The method of claim 1 , wherein broadcasting, by the first system controller, an alert to the plurality of both the second system controller and the plurality of node controllers comprises utilizing, by the first system controller, a functional subsystem interface (FSI) in the server computer to communicate with the plurality of system controllers and the plurality of node controllers.
5. The method of claim 1 , wherein determining, by the first system controller, that a simultaneous dump should be performed by a plurality of both the second system controller and the plurality of node controllers comprises utilizing a predetermined map that defines scenarios for which a simultaneous dump is required.