Computing system with reboot tracking

A method implemented at a computing system, the method comprising:

• reading out a stored reboot request identifier assigned to a node in the computing system, wherein the node is hosted on a host of the computing system, the stored reboot request identifier including a first value;
• receiving a first reboot request to reboot the node in the computing system, the first reboot request including a first reboot request identifier identifying the first reboot request;
• responsive to identifying a match between a value of the first reboot request identifier and the first value of the stored reboot request identifier, accepting the first reboot request and updating the stored reboot request identifier assigned to the node with a second value;
• receiving a second reboot request to reboot the node, the second reboot request including a second reboot request identifier identifying the second reboot request;
• responsive to identifying a mismatch between a value of the second reboot request identifier and the second value of the stored reboot request identifier, rejecting the second reboot request;
• and
• accepting or rejecting a request to change a state of the host based on whether the node is undergoing reboot.

A computing system, comprising:

• a processor and memory storing instructions that cause the processor to execute:
• a fabric layer configured to:
• initialize a tracked reboot state for a node in the computing system, the node being associated with a host in the computing system, the tracked reboot state including a stored reboot request identifier assigned to the node;
• read out the stored reboot request identifier assigned to the node;
• receive a reboot request to reboot the node in the computing system, the reboot request including a reboot request identifier identifying the reboot request;
• based at least on comparing the stored reboot request identifier assigned to the node with the reboot request identifier associated with the reboot request, determining whether to accept or reject the reboot request at a node level;
• and
• responsive to the reboot request being accepted at the node level and at a host level,
• add a reboot work item to a queue associated with the node, the reboot work item being configured to cause reboot of the node upon execution of the reboot work item, the queue including work items to be executed when the node is in a ready state;
• and
• assign a new reboot request identifier to the node,
• a host management layer configured to:
• based at least on the tracked reboot state, determine a state of the host;
• and
• based at least on the state of the host, determine whether to accept or reject the reboot request at the host level,
• a compute resource provider (CRP) layer configured to:
• based at least on the tracked reboot state, determine whether the reboot request was fulfilled and the node was successfully rebooted;
• and
• output a notification indicating whether the node was successfully rebooted.