EP-4736381-A1 - TECHNIQUES FOR ROTATING NETWORK ADDRESSES IN PREFAB REGIONS

EP4736381A1EP 4736381 A1EP4736381 A1EP 4736381A1EP-4736381-A1

Abstract

Techniques are disclosed for rotating network addresses following the installation of a prefab region network at a destination site. A manager service executing within a distributed computing system can allocate a rotation network address pool to a root allocator service that may be configured to provide network addresses from network address pools to dependent nodes within the distributed computing system, with each dependent node associated with a corresponding first network address of the network address pools. The manager service can receive an indication that a second network address of the rotation network address pool is associated with a dependent node. In response, the manager service can execute a migration operation for the dependent node to redirect network traffic within the distributed computing system from the first network address to the second network address.

Inventors

ADOGLA, EDEN
KUEHNEL, THOMAS
PALADINO, Zackery Vincent
DUFFEY, Laura Ann
RODRIGUEZ, Kimberly A.
GARAKAHALLI, NAVANEETH

Assignees

Oracle International Corporation

Dates

Publication Date: 20260506
Application Date: 20240625

Claims (20)

WHAT IS CLAIMED IS: 1. A computer-implemented method, comprising: allocating, by a manager service, a rotation network address pool to a root allocator service, the root allocator service configured to provide network addresses from network address pools to dependent nodes within a distributed computing system, the network address pools managed by the root allocator service, and each dependent node associated with a corresponding first network address of the network address pools; receiving, by the manager service, an indication that a second network address of the rotation network address pool is associated with a dependent node; and responsive to the indication, executing, by the manager service, a migration operation for the dependent node to redirect network traffic within the distributed computing system from the first network address to the second network address.
2. The computer-implemented method of claim 1, wherein the dependent node of the dependent nodes comprises a first networking interface associated with the first network address, and wherein executing the migration operation comprises: generating a second networking interface at the dependent node, the second networking interface associated with the second network address from the rotation network address pool; instructing the dependent nodes within the distributed computing system to direct networking traffic to the second networking interface; and deprecating the first networking interface of the dependent node.
3. The computer-implemented method of claim 2, wherein instructing the dependent nodes to direct networking traffic to the second networking interface comprises updating a domain name system (DNS) record to be associated with the second network address from the rotation network address pool.
4. The computer-implemented method of claim 1, wherein the dependent node of the dependent nodes is a first dependent node associated with the first network address, and wherein executing the migration operation comprises: implementing a second dependent node of the dependent nodes configured to provide the functionality of the first dependent node; associating the second network address from the rotation network address pool to the second dependent node; instructing the dependent nodes within the distributed computing system to direct networking traffic to the second network address of the second dependent node; and removing the first dependent node.
5. The computer-implemented method of claim 1, wherein the dependent node is a singleton node associated with the first network address, wherein the first network address is a swappable network address, and wherein executing the migration operation comprises replacing the first network address of the singleton node with the second network address from the rotation network address pool.
6. The computer-implemented method of claim 1, wherein the dependent node is a singleton node associated with the first network address, wherein the first network address is a fixed network address, and wherein executing the migration operation comprises: removing the singleton node; and replacing the singleton node with a new node associated with the second network address from the rotation network address pool.
7. The computer-implemented method of claim 1, wherein the dependent nodes of the distributed computing system comprise at least one first level allocator, and further comprising allocating a subset of the rotation network address pool by the first-level allocator to a subset of the dependent nodes.
8. The computer-implemented method of claim 1, further comprising: receiving, by the manager service, an additional indication that one or more computing devices of the distributed computing system were successfully connected to a network fabric at a destination site; and responsive to the indication, executing, by the manager service, the migration operation.
9. The computer-implemented method of claim 1, further comprising: receiving, by the manager service, an additional indication that an assigned pool of network addresses has been released to the root allocator service; and identifying the assigned pool of network addresses as available for an allocation.
10. A distributed computing system, comprising: one or more processors; and one or more memories storing computer-executable instructions that, when executed by the one or more processors, cause the distributed computing system to: allocate, by a manager service, a rotation network address pool to a root allocator service, the root allocator service configured to provide network addresses from network address pools to dependent nodes within a distributed computing system, the network address pools managed by the root allocator service, and each dependent node associated with a corresponding first network address of the network address pools; receive, by the manager service, an indication that a second network address of the rotation network address pool is associated with a dependent node; and responsive to the indication, execute, by the manager service, a migration operation for the dependent node to redirect network traffic within the distributed computing system from the first network address to the second network address.
11. The distributed computing system of claim 10, wherein the dependent node of the dependent nodes comprises a first networking interface associated with the first network address, and wherein executing the migration operation comprises: generating a second networking interface at the dependent node, the second networking interface associated with the second network address from the rotation network address pool; instructing the dependent nodes within the distributed computing system to direct networking traffic to the second networking interface; and deprecating the first networking interface of the dependent node.
12. The distributed computing system of claim 11, wherein instructing the dependent nodes to direct networking traffic to the second networking interface comprises updating a domain name system (DNS) record to be associated with the second network address from the rotation network address pool.
13. The distributed computing system of claim 10, wherein the dependent node of the dependent nodes is a first dependent node associated with the first network address, and wherein executing the migration operation comprises: executing a second dependent node of the dependent nodes configured to provide the functionality of the first dependent node; associating the second network address from the rotation network address pool to the second dependent node; instructing the dependent nodes within the distributed computing system to direct networking traffic to the second network address of the second dependent node; and removing the first dependent node.
14. The distributed computing system of claim 10, wherein the dependent node is a singleton node associated with the first network address, wherein the first network address is a swappable network address, and wherein executing the migration operation comprises replacing the first network address of the singleton node with the second network address from the rotation network address pool.
15. The distributed computing system of claim 10, wherein the dependent node is a singleton node associated with the first network address, wherein the first network address is a fixed network address, and wherein executing the migration operation comprises: removing the singleton node; and replacing the singleton node with a new node associated with the second network address from the rotation network address pool.
16. A non-transitory computer-readable medium storing computer- executable instructions that, when executed by one or more processors, cause a distributed computing system to: allocate, by a manager service, a rotation network address pool to a root allocator service, the root allocator service configured to provide network addresses from network address pools to dependent nodes within a distributed computing system, the network address pools managed by the root allocator service, and each dependent node associated with a corresponding first network address of the network address pools; receive, by the manager service, an indication that a second network address of the rotation network address pool is associated with a dependent node; and responsive to the indication, execute, by the manager service, a migration operation for the dependent node to redirect network traffic within the distributed computing system from the first network address to the second network address.
17. The non-transitory computer-readable medium of claim 16, wherein the dependent node of the dependent nodes comprises a first networking interface associated with the first network address, and wherein executing the migration operation comprises: generating a second networking interface at the dependent node, the second networking interface associated with the second network address from the rotation network address pool; instructing the dependent nodes within the distributed computing system to direct networking traffic to the second networking interface; and deprecating the first networking interface of the dependent node.
18. The non-transitory computer-readable medium of claim 17, wherein instructing the dependent nodes to direct networking traffic to the second networking interface comprises updating a domain name system (DNS) record to be associated with the second network address from the rotation network address pool.
19. The non-transitory computer-readable medium of claim 16, wherein the dependent node of the dependent nodes is a first dependent node associated with the first network address, and wherein executing the migration operation comprises: implementing a second dependent node of the dependent nodes configured to provide the functionality of the first dependent node; associating the second network address from the rotation network address pool to the second dependent node; instructing the dependent nodes within the distributed computing system to direct networking traffic to the second network address of the second dependent node; and removing the first dependent node.
20. The non-transitory computer-readable medium of claim 16, wherein the dependent node is a singleton node associated with the first network address, wherein the first network address is a swappable network address, and wherein executing the migration operation comprises replacing the first network address of the singleton node with the second network address from the rotation network address pool.

Description

PATENT ATTORNEY DOCKET NO.: 088325-1432134 (344100PC) Client Reference No.: ORC22134791-WO-PCT (IaaS #461) TECHNIQUES FOR ROTATING NETWORK ADDRESSES IN PREFAB REGIONS CROSS REFERENCES TO RELATED APPLICATIONS [0001] The present international application claims the benefit of and priority to U.S. Non- Provisional Application No. 18/215,632, filed on June 28, 2023, entitled "TECHNIQUES FOR ROTATING NETWORK ADDRESSES IN PREFAB REGIONS," Attorney Docket No. 088325-1307193 (344100US), the entire contents of which are herein incorporated by reference in their entirety for all purposes. FIELD [0002] This disclosure is related to cloud computing data centers. More particularly, this disclosure describes techniques for rotating network addresses for nodes within a data center network from network address pools available when constructing the data center network at a prefab factory to network address pools available for the data center network at a destination site. BACKGROUND [0003] A cloud infrastructure provider may operate one or more data centers in geographic areas around the world. A "region" is a logical abstraction around a collection of the computing, storage, and networking resources of the data centers of a given geographical area that are used to provide the cloud computing infrastructure. Building new regions can include provisioning the computing resources, configuring infrastructure, and deploying code to those resources, typically over network connections to the data centers. However, building regions with physical resources located at the final destination data center sites requires significant preparation work at the data centers that can complicate the logistics and scheduling of completing the building of a region. BRIEF SUMMARY [0004] Embodiments of the present disclosure relate to automatically building a region using a prefab factory. A prefab factory may be a facility dedicated to configuring computing devices, networking devices, and other physical resources for delivery to a destination site (e.g., a destination region—one or more data centers in a geographic area, a customer facility, etc.). Operations for building a region can include bootstrapping (e.g., provisioning and/or deploying) resources (e.g., infrastructure components, artifacts, etc.) for any suitable number of services available from the region when delivered to the destination. Once the physical resources have been configured at the prefab factory, they may be shipped to the destination site, installed at the destination data center, and have final configurations and other software resources deployed to the physical resources. Resources used for bootstrapping (e.g., software artifacts, software images, etc.) may be provided in a bootstrapping environment in an existing region (e.g., one or more data centers of a host region). The host region can be selected based on network proximity to the prefab factory, and in a complimentary fashion, the prefab factory may be sited to have high performance network connectivity to one or more host regions to support the bootstrapping environment. Building the region may be orchestrated by one or more cloud-based services that can manage the inventory of physical computing devices used to build regions in the prefab factory, generate and specify the configurations of regions to be built in the prefab factory, manage the bootstrapping of the regions, configure the regions for transmission to a destination site, and test and verify the physical resources after the physical resources have been installed at the destination site. A prefab region may be built to meet a specific customer’s configuration preferences (built-to- order) or built to a common specification that may be further customized during installation at a specific customer’s site (built-to-stock). [0005] One embodiment is directed to a computer-implemented method for rotating network addresses within a region network after or during installation at a destination site. The method can be performed by a manager service executing on one or more computing devices. The method can include allocating a rotation network address pool to a root allocator service. The root allocator service can be configured to provide network addresses from network address pools to dependent nodes within a distributed computing system. The network address pools can be managed by the root allocator service and each dependent node can be associated with a first network address of the network address pools. The method can also include receiving an indication that a second network address of the rotation network address pool is associated with a dependent node, and responsive to the indication, executing a migration operation for the dependent node to redirect network traffic within the distributed computing system from the first network address to the second network address. In some examples, redirecting traffic from the first network address to the second net