Scaling service capacity
AssigneeAmazon Technologies, Inc.
InventorsAbhinav A Agshikar, Ankur Goenka, Bhavya Gupta, Atul Agarwal, Manish Kumar Shukla
Life of this patent
Estimating the term…
Abstract
A scaling system is disclosed. The scaling system scales services for a high velocity event. To manage the scaling, a hierarchy of instruction routing network nodes is generated. The instruction routing network nodes provide scaling instructions to some of the services for the HVE. Some of the instruction routing network nodes generate or forward scaling instructions to down hierarchy instruction routing network nodes.
Claims20
A system, comprising:
- one or more processors;
- and
- one or more memories storing instructions that, upon execution by the one or more processors, configure:
- a first set of service components to perform a first service function;
- and
- a second set of service components to perform a second service function;
- a first instruction routing network node associated with the first service function;
- and
- a second instruction routing network node associated with the second service function,
- wherein the first instruction routing network node is configured to:
- receive event information characterizing a demand change event, the event information describing a demand or a change in demand in service capacity for the demand change event,
- generate first service component instructions based at least in part on the event information,
- transmit the first service component instructions to the first set of service components, the first service component instructions causing the first set of service components to change a service capacity of the first set of service components, generate first scaling instructions based at least in part on the event information, and
- transmit the first scaling instructions to the second instruction routing network node,
- wherein the second instruction routing network node is configured to:
- receive the first scaling instructions,
- generate second service component instructions based at least in part on the first scaling instructions, and
- transmit the second service component instructions to the second set of service components, the second service component instructions causing the second set of service components to change a service capacity of the second set of service components.
A computer-implemented method, comprising:
- receiving, with a first instruction routing network node, event information characterizing a demand change event, the event information describing a demand or a change in demand in service capacity for the demand change event, the first instruction routing network node being associated with a first service function;
- changing, with the first instruction routing network node, a service capacity of a first set of service components based at least in part on the event information;
- generating, with the first instruction routing network node, first scaling instructions based at least in part on the event information;
- transmitting, with the first instruction routing network node, the first scaling instructions to a second instruction routing network node, the second instruction routing network node being associated with a second service function;
- and
- changing, with the second instruction routing network node, a service capacity of a second set of service components based at least in part on the first scaling instructions.
A computer-readable storage medium storing computer-executable instructions that, when executed by one or more processors of a first instruction routing network node, cause the first instruction routing network node to:
- receive first scaling instructions associated with a demand change event;
- change a service capacity of a first set of service components based at least in part on the first scaling instructions;
- generate second scaling instructions based at least in part on the first scaling instructions;
- and
- transmit the second scaling instructions to a second instruction routing network node, the second scaling instructions causing the second instruction routing network node to change a service capacity of a second set of service components.
Dependencies are inferred from the claim text.