Advanced cloud architectures for power outage mitigation and flexible resource use

A method comprising:

• receiving a workload designated for execution on a non-uninterruptible power supply form (non-UPS) compatible computing resource with an indication, from a sender of the workload, to request execution of the workload using a non-UPS compatible computing resource, wherein a non-UPS compatible computing resource does not comply with uninterruptible power supply aspects of Network Equipment-Building System (NEBS) level 3;
• receiving a configuration for the workload, the configuration related to performance of the workload in an event of a power failure and cooling supply failure and the configuration comprising threshold temperature, performance criteria, and associated actions;
• and
• dispatching the workload on a first non-UPS compatible computing resource for execution by the first non-UPS compatible computing resource based on the indication from the sender of the workload to execute the workload using a non-UPS compatible computing resource, wherein in response to a failure of a power source to the first non-UPS compatible computing resource and an error rate of the workload indicating unacceptable performance based on the performance criteria, causing migration of the workload to a UPS compatible resource or another non-UPS compatible resource for execution.

A power management and resource management system comprising:

• a communication interface;
• at least one uninterruptible power supply form (UPS) compatible computing resource, wherein a UPS compliant computing resource is compatible with uninterruptible power supply aspects of Network Equipment-Building System (NEBS) level 3;
• at least one non-uninterruptible power supply form (non-UPS) compatible computing resource;
• and
• at least one processor communicatively coupled to the communication interface, the at least one UPS compatible computing resource, and the at least one non-UPS compatible computing resource, the at least one processor to:
• receive a workload identified for execution using a UPS compatible computing resource;
• cause execution of the workload using a first UPS compatible computing resource;
• receive a second workload and an indication from a sender of the workload that identifies the second workload is to be executed using a non-UPS compatible computing resource;
• receive a configuration for the second workload based on a power source failure or a cooling source failure, wherein the configuration comprises threshold temperature, performance criteria, and associated actions;
• cause execution of the second workload using a first non-UPS compatible computing resource based on the indication from the sender of the workload that identifies the second workload is to be executed using a non-UPS compatible computing resource, wherein in response to a failure of a power source to the first non-UPS compatible computing resource and an error rate of the second workload indicating unacceptable performance based on the performance criteria, the at least one processor is to cause migration of the second workload to a UPS compatible resource or another non-UPS compatible resource for execution;
• and
• cause results from the workload and the second workload to be transmitted to at least one workload requester.

A client device comprising:

• a network interface and
• at least one processor communicatively coupled to the network interface, wherein the at least one processor is to:
• discover, at a node, an uninterruptible power supply form (UPS) compatible computing resource and non-uninterruptible power supply form (non-UPS) compatible resource, wherein a UPS compatible computing resource is compatible with uninterruptible power supply aspects of Network Equipment-Building System (NEBS) level 3;
• cause a workload request to be transmitted to the node using the network interface, with an indication associated with the workload request to identify a workload associated with the workload request is to be performed by a non-UPS compatible computing resource;
• cause a configuration for the workload based on a power source failure or a cooling source failure to be transmitted to the node using the network interface, wherein the configuration comprises threshold temperature, performance criteria, and associated actions and wherein based a failure of a power source to the non-UPS compatible computing resource and an error rate of the workload indicating unacceptable performance based on the performance criteria, the workload is migrated for execution on a UPS compatible resource or another non-UPS compatible resource;
• and
• receive a result of execution of the workload from the node.

At least one non-transitory computer-readable medium comprising instructions stored thereon, that if executed by one or more processors, cause the one or more processors to:

• based on a reduction in power supplied to an edge server:
• allocate at least one workload for execution on an uninterruptible power supply form (UPS) compatible computing resource based on an indication that the at least one workload is to be executed using a UPS compatible computing resource;
• allocate at least one workload for execution on a non-uninterruptible power supply form (non-UPS) compatible computing resource based on an indication that the at least one workload is to be executed using a non-UPS compatible computing resource, wherein based a failure of a power source to the non-UPS compatible computing resource and an error rate of the at least one workload indicating unacceptable performance based on performance criteria, migrate the at least one workload for execution on a UPS compatible resource or another non-UPS compatible resource;
• and
• output a result of execution of the at least one workload from execution by the UPS compatible computing resource and a result of execution of the at least one workload from execution by the non-UPS compatible computing resource, wherein the edge server comprises the UPS compatible computing resource and the non-UPS compatible computing resource.