Dynamic provisioning for data replication groups

• US 10,733,201 B1
• Filed: 11/30/2015
• Issued: 08/04/2020
• Est. Priority Date: 11/30/2015
• Status: Active Grant

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Inventors: Chen, Tao, Parthasarathy, Madhuvanesh, Agrawal, Shobha, Ping, Fan, Jain, Divya Ashok Kumar, Brooker, Marc John, Zulak, Matthew Joseph, Pallikila, Lakshmi N.
• Primary Examiner(s): Trujillo, James
• Assistant Examiner(s): Morris, John J
• Application Number:
  US14/954,987
• Time in Patent Office: 1,709 Days
• Field of Search: 707623
• CPC Class Codes: G06F 11/1402 Saving, restoring, recovering or retrying
• CPC Class Codes: G06F 11/2056 by mirroring
• CPC Class Codes: G06F 16/27 Replication, distribution or synchronisation of data between databases or within a distributed database system; Distributed database system architectures therefor
• CPC Class Codes: H04L 67/1095 Replication or mirroring of data, e.g. scheduling or transport for data synchronisation between network nodes

Abstract

Data replication groups may be used to store data in a distributed computing environment. The data replication groups may include a set of nodes executing a consensus protocol to maintain data durably. In order to increase the speed of provisioning, a data replication may be provision with less than the total number of nodes to be included in the group. Once an initial set of nodes have been provisioned and are implementing a consensus protocol, the data replication group may begin to receive traffic. Furthermore, new nodes may be added to the group until the total number of nodes for the group has been reached.

Claims

1. A computer-implemented method, comprising:

receiving a request for a data replication group, the data replication group to have a total number of nodes;

fulfilling the request by at least;

provisioning a first number of nodes of the data replication group, the first number of nodes being less than the total number of nodes, wherein the first number of nodes is selected according to a latency threshold indicated in the request;

implementing a consensus protocol on the first number of nodes of the data replication group;

storing information separate from consensus protocol data, the information corresponding to the data replication group indicating that the first number of nodes of the data replication group is ready to receive traffic, the information configured to enable the one or more computer systems to determine that the indicated first number of nodes of the data replication group can be used to fulfill requests; and

mapping location information of the data replication group with a requestor responsible for submitting the request; and

adding new nodes to the data replication group until the total number of nodes of the data replication group is reached while the data replication group receives traffic on behalf of the requestor.

2. The computer-implemented method of claim 1, wherein the first number of nodes of the data replication group is one.

3. The computer-implemented method of claim 1, wherein adding new nodes to the data replication group further comprises adding a plurality of new nodes to the group at a certain point in time.

4. The computer-implemented method of claim 1, wherein adding new nodes to the data replication group further comprises identifying a group of server computer systems for hosting the new nodes based at least in part on one or more constraints on placement of the new nodes in a distributed computing environment, the one or more constraints on placement causing the group of computer systems to be identified such that at least a first portion of the node are hosted by a first computer system of the group of server computer system and a second portion of the nodes are hosted by a second computer system of the group of server computer systems.

5. A system, comprising:

one or more processors;

memory that includes instructions that, when executed by the one or more processors, cause the system to;

provision a data replication group with a first number of nodes, where the first number of nodes is less than a total number of nodes of the data replication group, wherein the first number of nodes is selected based at least in part on a provisioning time associated with the data replication group;

while the data replication group has the first number of nodes but before having the total number of nodes, associate the data replication group with a computing resource such that the data replication group stores data on behalf of the computing resource;

provision a second number of nodes over an interval of time; and

add the provisioned second number of nodes to the data replication group, where the first number of nodes combined with the second number of nodes is equal to or greater than the total number of nodes of the data replication group.

6. The system of claim 5, wherein the memory further includes instructions that, when executed by the one or more processors, cause the system to provision the data replication group as a result of receiving a request, from a requestor responsible for operating the computing resource, for the data replication group.

7. The system of claim 5, wherein the memory further includes instructions that, when executed by the one or more processors, cause the system to include the data replication group in a pool of data replication groups.

8. The system of claim 5, wherein the memory further includes instructions that, when executed by the one or more processors, cause the system to cause the first number of nodes to implement a consensus protocol, where the consensus protocol does not require all members implementing the protocol to be initialized before the protocol can be executed.

9. The system of claim 8, wherein the consensus protocol is a Paxos consensus protocol.

10. The system of claim 5, wherein the memory further includes instructions that, when executed by the one or more processors, cause the system to identify one or more server computer systems for placement of the second number of nodes.

11. The system of claim 5, wherein the memory further includes instructions that, when executed by the one or more processors, cause the system to:

receive, prior to provisioning the second number of nodes, a request to store data; and

cause the data replication group including the first number of nodes to fulfill the request.

12. The system of claim 5, wherein the memory further includes instructions that, when executed by the one or more processors, cause the system to replace one or more unhealthy nodes of the data replication group with one or more new nodes.

13. A non-transitory computer-readable storage medium having stored thereon executable instructions that, when executed by one or more processors of a computer system, cause the computer system to at least:

cause a first number of nodes of a data replication group to be executed by a host computer system, where the data replication group comprises a total number of nodes greater than the first number of nodes, wherein the first number of nodes is selected to reduce a provisioning time associated with the data replication group;

enable the data replication group to receive traffic while comprising the first number of nodes but before comprising the total number of nodes;

cause a set of new nodes to be executed by one or more host computer systems, where the first number of nodes and the set of nodes are greater than or equal to the total number of nodes of the data replication group; and

cause the set of new nodes to be added to the data replication group.

14. The non-transitory computer-readable storage medium of claim 13, wherein the instructions further comprise instructions that, when executed by the one or more processors, cause the computer system to implement a consensus protocol between one or more nodes of the data replication group, where the consensus protocol does not require the one or more nodes to be initialized before the protocol can be executed.

15. The non-transitory computer-readable storage medium of claim 13, wherein the instructions further comprise instructions that, when executed by the one or more processors, cause the computer system to store customer data among one or more nodes of the data replication group.

16. The non-transitory computer-readable storage medium of claim 13, wherein the instructions further comprise instructions that, when executed by the one or more processors, cause the computer system to store data received from one or more services of a computing resource service provider.

17. The non-transitory computer-readable storage medium of claim 13, wherein the instructions that cause the computer system to cause a set of new nodes to be executed by one or more host computer systems further include instructions that cause the computer system to identify the one or more host computer systems by at least transmitting a request to a service of a computing resource service provider for potential placement locations of the nodes on the one or more host computer system based at least in part on one or more constraints.

18. The non-transitory computer-readable storage medium of claim 13, wherein the instructions further comprise instructions that, when executed by the one or more processors, cause the computer system to replicate data stored by at least one of the first number of nodes to the new nodes of the data replication group.

19. The non-transitory computer-readable storage medium of claim 13, wherein the instructions further comprise instructions that, when executed by the one or more processors, cause the computer system to add the new nodes to the data replication group individually.

20. The non-transitory computer-readable storage medium of claim 13, wherein the instructions further comprise instructions that, when executed by the one or more processors, cause the computer system to concurrently add a plurality of the new nodes to the data replication group over an interval of time.