Abstract: In one embodiment, a method receives a software command from a software element in a main computer at a management computer via an application programming interface (API) included in the management computer. The management computer manages the main computer, and the main computer controls services for a distributed computing system. The management computer then determines a hardware device associated with the software command where the management computer includes a communication pathway to the hardware device. The management computer translates the software command to a hardware command that the hardware device can process where the hardware device cannot process the software command. Then, the management computer sends the hardware command to the hardware device via the communication pathway where the management computer provides the API to allow the software element to communicate with the hardware element without the software element having to translate the software command to the hardware command.

Abstract: In one embodiment, a method receives a software upgrade package for a management computer and main computer. The management computer upgrades software of the management computer using the software upgrade package where the upgrading replaces an image of the software of the management computer with an image from the software upgrade package. Upon upgrade of the management computer, the management computer initiates an upgrade of the main computer. The main computer withdraws use of the services, and upon the withdrawing, the management computer reboots the main computer. Then, the main computer upgrades software of the main computer using the software upgrade package upon rebooting where the upgrading replaces an image of the software of the main computer with an image from the software upgrade package. Upon the upgrading, the main computer restores the use of the services.

Abstract: In one embodiment, a method detects a failure of a container in a controller node where the container includes a service being performed and isolated from other services being performed in other containers on the controller node. The controller node terminates the container including the service and determines a known state for the service. The known state is known to be operational without including a cause of the failure and the service operated from the known state saving changes to the known state during operation separately from the known state. The controller node restarts the service in a new container that replaces the terminated container where the restarted service starts from the known state without using the changes.

Abstract: In one embodiment, a method receives a software upgrade package for a management computer and main computer. The management computer upgrades software of the management computer using the software upgrade package where the upgrading replaces an image of the software of the management computer with an image from the software upgrade package. Upon upgrade of the management computer, the management computer initiates an upgrade of the main computer. The main computer withdraws use of the services, and upon the withdrawing, the management computer reboots the main computer. Then, the main computer upgrades software of the main computer using the software upgrade package upon rebooting where the upgrading replaces an image of the software of the main computer with an image from the software upgrade package. Upon the upgrading, the main computer restores the use of the services.