PROVIDING BOOT DATA IN A CLUSTER NETWORK ENVIRONMENT

摘要

A computer cluster includes a group of connected computers that work together essentially as a single system. Each computer in the cluster is called a node. Each node has a boot device configured to load an image of an operating system into the node's main memory. Sometimes the boot device of a first node experiences a problem that prevents the operating system from loading. This can affect the entire cluster. Some aspects of the disclosure, however, are directed to operations that determine the problem with the first node's boot device based on a communication sent via a first communications network. Further, the operations can communicate to the first node a copy of boot data from a second node's boot device. The copy of the boot data is sent via a second communications network different from the first communications network. The copy of the boot data can solve the first boot device's problem.

主权项

1. A method comprising:
determining, by a first node computing device, when there is a problem with a boot device of a second node computing device based, at least in part, on a communication received from the second node computing device, wherein the communication is received via a first network that is different than a second network used for data transfers among nodes of a cluster; determining, by the first node computing device, boot data that can be used to resolve the problem when the determining indicates the problem with the boot device of the second node computing device; selecting, by the first node computing device and from another boot device, at least a portion of the boot data that can be used to resolve the problem; and providing, by the first node computing device, to the second node computing device, and via the second network, a copy of the at least the portion of the boot data selected from the another boot device.