Datacenter event handling

Patent metadata

Field	Value
Publication number	US-9594721-B1
Application number	US-201213693793-A
Country	US
Kind code	B1
Filing date	Dec 4, 2012
Priority date	Dec 4, 2012
Publication date	Mar 14, 2017
Grant date	Mar 14, 2017

Abstract

Systems and methods for handling resources in a computer system differently in certain situations, such as catastrophic events, based upon an assigned layer of the resource to the system. The layer can be based, for example, on criticality of the resource to the system. Services or computing device resources can be physically segregated in accordance with layers and can be managed in accordance with the segregation. In response to receiving information about an event, the different layers can be handled in accordance with their criticality, for example by shutting some of the resources down and/or slowing some of the resources down.

First claim

What is claimed is:

1. A computer-implemented method for managing datacenter components in a datacenter, comprising:
under the control of one or more computer systems configured with executable instructions:
receiving, from a client device, an indication of a specified datacenter component and a criticality level of the datacenter component, the criticality level being associated with an assigned layer;
maintaining information about a plurality of datacenter components in a datacenter that includes the specified datacenter component and a plurality of layers, the information including, for each of the plurality of datacenter components, an assigned one of the plurality of layers for the datacenter component, the specified datacenter component being assigned to the assigned layer;
accessing layer information for each layer of the plurality of layers, the layer information including information on usage of a resource for the layer of the plurality of layers, the usage of the resource for the layer being determined from operational characteristics of each datacenter component, the operational characteristics provided by each datacenter component assigned to the layer of the plurality of layers;
receiving information about an event related to the datacenter occurring independent of the plurality of datacenter components, the event affecting an amount of the resource available;
in response to receiving the information about the event, determining, based at least in part on the information on usage of the resource in the accessed layer information associated with at least one layer, a change to operation of datacenter components assigned to the at least one layer such that usage of the resource by datacenter components of the plurality of layers does not exceed the total amount of the resource available; and
changing operation of the determined at least one layer of the plurality of layers.

2. The computer-implemented method of claim 1 , wherein changing operation comprises changing a processor state associated with the datacenter components.

3. The computer-implemented method of claim 2 , wherein the processor state comprises at least one of processor performance state (P-state), processor operating state (C-state), or processor sleep state (S-state).

4. The computer-implemented method of claim 1 , wherein changing operation comprises changing at least one of a device or global state associated with the datacenter components of the at least one layer of the plurality of layers.

5. The computer-implemented method of claim 1 , wherein changing operation comprises shutting down the datacenter components of the at least one layer of the plurality of layers.

6. The computer-implemented method of claim 1 , wherein changing operation comprises slowing down operation of the datacenter components of the at least one layer of the plurality of layers.

7. A computer-implemented method for managing components in a computing environment, comprising:
under the control of one or more computer systems configured with executable instructions,
receiving an indication of at least one computer component and a criticality level of the at least one computer component; and
in response to receiving information about an event in a computing environment affecting an amount of a common resource available, the computing environment comprising a network of a plurality of computer components sharing the common resource that includes the at least one computer component:
accessing information about the plurality of computer components in the computing environment and a plurality of layers, the information including, for each of the plurality of computer components, an assigned one of the plurality of layers for the computer component, the one of the plurality of layers having been assigned based at least in part on a criticality of the computer components to the computing environment;
determining a resource usage of the common resource for each layer of the plurality of layers from operational characteristics of each datacenter component, the operational characteristics provided by each datacenter component assigned to the layer of the plurality of layers;
determining at least one critical layer of the plurality of layers based at least in part on the criticality of the computer components in the at least one critical layer, the at least one critical layer determined such that the resource usage of the computer components in said at least one critical layer is below the amount of the common resource available; and
causing operation to change for at least some of the plurality of computer components in accordance with the assigned layer for the computer component, the at least some of the plurality of computer components being assigned to layers other than the at least one critical layer.

8. The computer-implemented method of claim 7 , wherein causing operation to change comprises reducing power consumption of said at least some of the plurality of computer components.

9. The computer-implemented method of claim 7 , wherein causing operation to change comprises changing operation of computer components assigned to one and only one layer of the plurality of layers in response to the event.

10. The computer-implemented method of claim 7 , wherein causing operation to change comprises changing a processor state associated with the computer components.

11. The computer-implemented method of claim 7 , wherein causing operation to change comprises changing computer components assigned a first layer to a first altered state and changing computer components assigned a second layer to a second altered state that is different than the first altered state.

12. The computer-implemented method of claim 7 , wherein causing operation to change comprises at least one of shutting down, slowing, or putting to sleep the computer components.

13. A datacenter environment, comprising:
a plurality of datacenter components;
a data store for maintaining information about the plurality of datacenter components and a plurality of layers, the information including, for each of the plurality of datacenter components, an assigned one of the plurality of layers for the datacenter component selected by a third party and operational characteristics of each of the plurality of datacenter components provided by the datacenter components; and
a management component, comprising:
at least one memory that stores computer-executable instructions; and
at least one processor configured to access the at least one memory, wherein the at least one processor is configured to execute the computer-executable instructions to collectively at least:
in response to receiving information about an event in the datacenter environment affecting an amount of a resource available to the datacenter components, access the information to identify at least one critical layer of the plurality of layers sustainable by the amount of the resource available to the datacenter components based on the operational characteristics of each of the plurality of datacenter components assigned to the at least one critical layer; and
cause operation to change for at least some of the plurality of datacenter components, the at least some of the plurality of datacenter components being those not assigned to the at least one critical layer.

14. The datacenter environment of claim 13 , wherein causing operation to change for at least some of the plurality of datacenter components comprises changing a processor state associated with the datacenter components.

15. The datacenter environment of claim 14 , wherein the pro

Assignees

• Amazon Tech Inc

Inventors

• Marr Michael David
• Bryan David Edward
• Wishman Max Jesse

Classifications

• G06F1/206

  comprising thermal management · CPC title

• G06F1/324

  by lowering clock frequency · CPC title

• G06F1/3296

  by lowering the supply or operating voltage · CPC title

• H05K7/1498

  Resource management, Optimisation arrangements, e.g. configuration, identification, tracking, physical location (thermal management H05K7/20836) · CPC title

• H05K7/1495

  providing data protection in case of earthquakes, floods, storms, nuclear explosions, intrusions, fire · CPC title