Multi-core re-initialization failure control system
A method of a computer system recovering from a core re-initialization failure is described. The method may include automatically detect a core re-initialization failure during a core re-initialization process by a hypervisor. The hypervisor automatically determines whether the core re-initialization failure is a permanent failure. If the core re-initialization failure is a permanent failure, then automatically determine, by the hypervisor, which cores are re-initialized and which cores are indeterminate. Automatically allocate the re-initialized cores between one or more virtual machines by the hypervisor.
1. A method of controlling a core re-initialization failure of a multi-core processor comprising:
automatically detecting a core re-initialization failure during a core re-initialization process by a hypervisor;
automatically determining whether the core re-initialization failure is a permanent failure by the hypervisor;
if the core re-initialization failure is a permanent failure, automatically determining by the hypervisor which cores are re-initialized and which cores are indeterminate; and
automatically allocating the re-initialized cores between one or more virtual machines by the hypervisor.
2. The method of claim 1 , wherein automatically allocating the re-initialized cores between one or more virtual machines by the hypervisor includes:
at the hypervisor, automatically determining a re-initialized core capacity;
at the hypervisor, automatically determining a capacity of a first virtual machine having a virtual machine capacity guarantee; and
at the hypervisor, automatically reducing a capacity of the one or more virtual machines with a low availability priority if the re-initialized core capacity does not meet the virtual machine capacity guarantee of the first virtual machine.
3. The method of claim 2 , further comprising:
at the hypervisor, automatically terminating one or more virtual machines to meet the virtual machine capacity guarantees of the first virtual machine.
4. The method of claim 3 , wherein the terminating one or more virtual machines includes terminating one or more virtual machines to open up enough core resources to meet the virtual machine capacity guarantees of the greatest number of virtual machines.
5. The method of claim 1 , wherein automatically determining by the hypervisor which cores are re-initialized and which cores are indeterminate includes determining whether the permanent failure is related to a re-initialization logic element, if the permanent error is related to the re-initialization logic element, then the core being re-initialized and remaining uninitialized cores are considered indeterminate, if the permanent failure is not related to the re-initialization logic element and related to the core being re-initialized, then the core being re-initialized may be considered indeterminate.
6. The method of claim 5 , wherein the re-initialization logic element is a power-on reset (POR) engine.
7. The method of claim 1 , further comprising:
at the hypervisor, automatically re-starting the core re-initialization process to re-initialize the uninitialized cores if the core re-initialization failure is not a permanent failure or if one or more cores are the cause of the permanent failure.