Automated data center selection

• US 10,142,406 B2
• Filed: 03/11/2013
• Issued: 11/27/2018
• Est. Priority Date: 03/11/2013
• Status: Active Grant

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Inventors: Suryanarayanan, Deepak, Brown, David Everard, Ah Kun, Malcolm Russell, Farrell, Eugene Michael
• Primary Examiner(s): Joshi, Suraj
• Assistant Examiner(s): Means, Jaren M
• Application Number:
  US13/794,595
• Publication Number:
  US 20140258374A1
• Time in Patent Office: 2,087 Days
• CPC Class Codes: H04L 67/101 based on network conditions
• CPC Class Codes: H04L 67/1021 based on client or server locations
• CPC Class Codes: H04L 67/1023 based on a hash applied to IP addresses or costs

Abstract

Systems and methods are presented for automatically selecting a data center that can communicate with a user computing system over a connection that provides reduced or minimal latency. In some cases, the system can select the data center based on the calculation of one or more latency factors that correlate, at least in part, to the latency of a communication channel between the data center and the user computing system. Further, the selected data center may be identified from a set of data centers that include one or more computing resources the user has identified as desiring to access.

Claims

1. A method for selecting a data center of a Program Execution Service (PES) platform, the method comprising:

under control of a PES platform that includes a number of data centers, at least some of the number of data centers capable of providing a user with access to one or more computing resources hosted by said at least some of the number of data centers, at least some of the computing resources including compute nodes comprising virtual desktops that enable the user to access a plurality of services provided by the PES platform;

receiving a request at the PES platform to obtain access to a computing resource from a computing device of the user;

determining, from the number of data centers, a set of data centers that include the computing resource;

identifying a plurality of data centers from the set of data centers based, at least in part, on a latency factor of each of the plurality of data centers;

selecting a data center from the plurality of data centers using a tie breaking factor, the tie breaking factor based at least in part on a utilization rate for the data centers of the plurality of data centers;

accessing access rules associated with an entity, the access rules specifying rules for distributing computing resource access of different users of a group of users at the entity among different instances of the computing resource, the group of users one of multiple groups of users, and the group of users including the user, wherein the access rules specify for a first group of users from the multiple groups of users;

     1) an amount of users permitted to access a first instance of the computing resource;

     2) that no more than a first amount of users from the first group of users share physical computing resources; and

     3) that users exceeding the first amount of users from the first group of users be assigned to a second instance of the computing resource, wherein each group of users is associated with a different role at the entity thereby reducing a number of users associated with a particular role that lose access to the computing resource when access to the first instance of the computing resource is lost, and wherein at least one user from a second group of users from the multiple groups of users is permitted to access the first instance of the computing resource;

determining whether granting the user of the computing device access to the computing resource at the data center will violate the access rules;

in response to determining that the access rules will be violated, selecting an alternative data center from the plurality of data centers, and

in response to determining that the access rules will not be violated by granting the user of the computing device access to the computing resource at the data center, granting the user of the computing device access to the computing resource at the data center.

2. The method of claim 1, further comprising determining the latency factor for each data center by:

calculating a distance between the computing device and the data center based on a geographic location of the computing device and a geographic location of the data center;

determining a first latency factor based on the distance between the computing device and the data center;

performing one or more latency tests between the computing device and the data center to determine a second latency factor;

identifying a network hop closest to the computing device;

retrieving historical latency information between the data center and the network hop; and

calculating the latency factor based, at least in part, on the first latency factor, the second latency factor, and the historical latency information.

3. The method of claim 2, wherein the one or more latency tests include one or more of the following:

a ping operation;

a traceroute operation;

a traceroute6 operation;

a tracert operation; and

a tracepath operation.

4. The method of claim 1, wherein determining the set of data centers that include the compute node comprises:

identifying data centers within a threshold distance of the geographic location of the computing device; and

identifying the set of data centers that include the compute node from the data centers identified within the threshold distance of the geographic location of the computing device.

5. A system for selecting a data center of a Program Execution Service (PES) platform comprising a plurality of hardware processors and non-volatile memory systems, the system comprising:

a PES platform comprising a number of data centers including a number of computing systems comprising one or more hardware processors and a memory, the PES platform receiving a request to obtain access to a computing resource from a computing device of a user;

a resource allocation system that;

determines from the number of data centers a set of data centers that include the computing resource;

identifies a plurality of data centers from the set of data centers based, at least in part, on a latency factor for at least some of the plurality of data centers; and

selects a data center from the plurality of data centers based at least in part on a utilization rate associated with the plurality of data centers and access rules for accessing the plurality of data centers, the access rules associated with a group of users at an entity, the group of users one of multiple groups of users, and the access rules specifying for a first group of users from the multiple groups of users;

     1) an amount of users permitted to access an instance of the computing resource during a time period;

     2) that no more than a first amount of users from the first group of users share physical computing resources; and

     3) that users exceeding the first amount of users from the first group of users be assigned to a different instance of the computing resource during the time period, wherein each group of users is associated with a different role at the entity thereby reducing a number of users associated with a particular role that lose access to the computing resource when access to the first instance of the computing resource is lost, and wherein at least one user from a second group of users from the multiple groups of users is permitted to access the first instance of the computing resource; and

a data center management system that grants the user of the computing device access to the computing resource at the data center.

6. The system of claim 5, further comprising a latency calculation system that calculates the latency factor by:

calculating a distance between the computing device and the data center based on the geographic location of the computing device and a geographic location of the data center; and

determining the latency factor based, at least in part, on the distance between the computing device and the data center.

7. The system of claim 5, further comprising a latency calculation system that calculates the latency factor by:

performing one or more latency tests between the computing device and the data center; and

determining the latency factor based, at least in part, on the results of the one or more latency tests.

8. The system of claim 5, further comprising a latency calculation system that calculates the latency factor by:

identifying a network hop closest to the computing device;

retrieving historical latency information between the data center and the network hop; and

determining the latency factor based, at least in part, on the historical latency information.

9. The system of claim 5, further comprising a latency calculation system that calculates the latency factor by:

calculating a distance between the computing device and the data center based on the geographic location of the computing device and a geographic location of the data center;

determining a first latency factor based on the distance between the computing device and the data center;

performing one or more latency tests between the computing device and the data center to determine a second latency factor;

identifying a network hop closest to the computing device;

retrieving historical latency information between the data center and the network hop; and

calculating the latency factor based, at least in part, on the first latency factor, the second latency factor, and the historical latency information.

10. The system of claim 9, wherein calculating the latency factor further comprises weighting one or more of the first latency factor, the second latency factor, or the historical latency information.

11. The system of claim 5, wherein the resource allocation system further identifies the set of data centers by:

identifying data centers within a threshold distance of a geographic location of the computing device; and

identifying the set of data centers that include the computing resource from the data centers identified within the threshold distance of the geographic location of the computing device.

12. The system of claim 5, wherein the resource allocation system further identifies the set of data centers from data centers that the user is authorized to access.

13. The system of claim 5, wherein the latency factor correlates, at least in part, to a communication latency between the computing device and the data center.

14. Non-transitory physical computer storage comprising computer-executable instructions that direct a computing system to perform a method for selecting a data center of a Program Execution Service (PES) platform, the method comprising:

receiving a request at a PES platform to obtain access to a computing resource from a computing device of a user, the PES platform comprising a number of data centers;

determining, from the number of data centers, a plurality of data centers that include the computing resource;

selecting a data center from the plurality of data centers based at least in part on a utilization rate associated with the plurality of data centers and access rules for accessing the plurality of data centers, the access rules associated with a group of users at an entity, the group of users one of multiple groups of users, and the access rules specifying for a first group of users from the multiple groups of users;

     1) an amount of users from the group of users permitted to access an instance of the computing resource during a time period;

     2) that no more than a first subset of users from the first group of users share physical computing resources; and

     3) that users from a second subset of users from the first group of users be assigned to a different instance of the computing resource during the time period, wherein each group of users is associated with a different role at the entity thereby reducing a number of users associated with a particular role that lose access to the computing resource when access to the first instance of the computing resource is lost, and wherein at least one user from a second group of users from the multiple groups of users is permitted to access the first instance of the computing resource; and

granting the user of the computing device access to the computing resource at the data center.

15. The non-transitory physical computer storage of claim 14, further comprising calculating a latency factor for each data center by:

calculating a distance between the computing device and the data center based on the geographic location of the computing device and a geographic location of the data center;

determining a first latency factor based on the distance between the computing device and the data center;

performing one or more latency tests between the computing device and the data center to determine a second latency factor;

identifying a network hop closest to the computing device;

retrieving historical latency information between the data center and the network hop; and

calculating the latency factor based, at least in part, on the first latency factor, the second latency factor, and the historical latency information.

16. The non-transitory physical computer storage of claim 15, wherein the one or more latency tests include one or more of the following:

a ping operation;

a traceroute operation;

a traceroute6 operation;

a tracert operation; and

a tracepath operation.

17. The non-transitory physical computer storage of claim 15, wherein identifying a network hop closest to the computing device comprises accessing a lookup table to identify a public network hop closest to the computing device.

18. The non-transitory physical computer storage of claim 15, wherein calculating the latency factor further comprises weighting one or more of the first latency factor, the second latency factor, or the historical latency information.

19. The non-transitory physical computer storage of claim 14, wherein determining the plurality of data centers that include the computing resource comprises:

identifying data centers within a threshold distance of a geographic location of the computing device; and

identifying the plurality of data centers that include the computing resource from the data centers identified within the threshold distance of the geographic location of the computing device.

20. The non-transitory physical computer storage of claim 14, wherein determining the plurality of data centers that include the computing resource comprises identifying data centers capable of obtaining access to the computing resource.

21. The non-transitory physical computer storage of claim 14, wherein determining the plurality of data centers that include the computing resource further comprises identifying the plurality of data centers from data centers that the user is authorized to access.

22. The non-transitory physical computer storage of claim 14, wherein the computing resource comprises a virtual desktop that enables the user to access a plurality of services provided by the PES platform.

23. The non-transitory physical computer storage of claim 22, wherein the plurality of services comprises one or more of the following:

an application;

a file manager;

or a file storage.

24. The non-transitory physical computer storage of claim 14, wherein the computing resource comprises an application.

25. The non-transitory physical computer storage of claim 15, wherein the latency factor correlates, at least in part, to a communication latency between the computing device and the data center.