Smart load balancing of containers for data protection using supervised learning

Patent metadata

Field	Value
Publication number	US-2025021446-A1
Application number	US-202318353103-A
Country	US
Kind code	A1
Filing date	Jul 16, 2023
Priority date	Jul 16, 2023
Publication date	Jan 16, 2025
Grant date	—

Abstract

A container load balancer process helps schedule backups of containerized data based on defined attributes and historical data. Containers are classified using a KNN-based classifier based on attributes. A tagger component assigns a priority tag to each container. A monitor monitors an assignment of backup tasks to proxies for the backing up step, and a load balancer determines if the assignment distributes backup loads within a defined performance tolerance, and adjusts the assignment if not. A backup server then backs up the container data in an order determined by the classifying and the assignment or adjusted assignment.

First claim

What is claimed is:

1 . A computer-implemented method of balancing backups and data processing loads of container data in a network, comprising:
classifying, with respect to a backup priority, each container of a plurality of containers storing the container data, and based on attributes of each container; generating a priority score for each container based on the classifying; tagging each container with a priority tag based on a priority score generated for each container based on the classifying; monitoring an assignment of backup tasks to proxies for the backing up step; determining if the assignment distributes backup loads within a defined performance tolerance, and adjusting the assignment if not; and backing up the container data in an order determined by the classifying and the assignment or adjusted assignment.

2 . The method of claim 1 wherein the container data is managed by a container management process, and wherein the network comprises a Kubernetes cluster having a controller, an application program interface (API) server, and a data process performs the backups.

3 . The method of claim 2 wherein each container is embodied as a Kubernetes Docker component and is deployed as portable, self-sufficient container.

4 . The method of claim 3 wherein the priority tag comprises metadata appended to the dataset, and wherein the tag is implemented as an alphanumeric string appended to an existing payload of the dataset having data source information.

5 . The method of claim 4 wherein the tag comprises a key value having a format of “PRIORITY_TAG” with an associated scalar value, wherein the associated value represents a priority value of a corresponding container ranked along a defined scale, and wherein the defined scale is on the order of 1 to 5, in one of ascending or descending order of priority.

6 . The method of claim 5 wherein the tag is associated with a Kubernetes payload of the dataset using a Kubernetes application programming interface (API).

7 . The method of claim 1 wherein the classifying comprises a k-nearest neighbors (KNN) algorithm that determines the backup priority based on a criticality of data in a container based on the attributes.

8 . The method of claim 7 further comprising training a model for the KNN algorithm using historical data of containers from users and laboratory environments to establish past priorities of backups of container data.

9 . The method of claim 8 wherein the classifying utilizes an artificial intelligence (AI) based component comprising a data collection component, a training component, and an inference component, and contains historical information regarding containers of the network to continuously train a machine learning (ML) algorithm to identify backup prioritization of container data.

10 . The method of claim 9 wherein the network comprises a PowerProtect Data Domain deduplication backup system.

11 . The method of claim 1 wherein the attributes comprise at least one of: container size, ownership, creation time, location, applications, datastore size, and provision type.

12 . A system for dynamically balancing backups and data processing loads of container data in a network, comprising:
a classifier, classifying each container of a plurality of containers storing the container data with respect to a backup priority, and based on attributes of each container; a component generating a priority score for each container based on the classifying; a tagger tagging each container with a priority tag based on the generated priority score; a monitor monitoring an assignment of backup tasks to proxies for the backing up step; a load balancer determining if the assignment distributes backup loads within a defined performance tolerance, and adjusting the assignment if not; and a backup server backing up the container data in an order determined by the classifying and the assignment or adjusted assignment.

13 . The system of claim 12 wherein the container data is managed by a container management process, and wherein the network comprises a Kubernetes cluster having a controller, an application program interface (API) server, and a data process performs the backups, and wherein each container is embodied as a Kubernetes Docker component and is deployed as portable, self-sufficient container.

14 . The system of claim 13 wherein the priority tag comprises metadata appended to the dataset, and wherein the tag is implemented as an alphanumeric string appended to an existing payload of the dataset having data source information, and further wherein the tag comprises a key value having a format of “PRIORITY_TAG” with an associated scalar value, wherein the associated value represents a priority value of a corresponding container ranked along a defined scale, and wherein the defined scale is on the order of 1 to 5, in one of ascending or descending order of priority.

15 . The system of claim 14 wherein the tag is associated with a Kubernetes payload of the dataset using a Kubernetes application programming interface (API).

16 . The system of claim 12 wherein the classifying comprises a k-nearest neighbors (KNN) algorithm that determines the backup priority based on a criticality of data in a container based on the attributes.

17 . The system of claim 16 further comprising a training component training a model for the KNN algorithm using historical data of containers from users and laboratory environments to establish past priorities of backups of container data.

18 . The system of claim 17 wherein the classifier comprises an artificial intelligence (AI) based component comprising a data collection component, a training component, and an inference component, and contains historical information regarding containers of the network to continuously train a machine learning (ML) algorithm to identify backup prioritization of container data.

19 . The system of claim 12 wherein the attributes comprise at least one of: container size, ownership, creation time, location, applications, datastore size, and provision type.

20 . A tangible computer program product having stored thereon program instructions that, when executed by a process, cause the processor to perform a method of balancing backups and data processing loads of container data in a network, comprising:
classifying, with respect to a backup priority, each container of a plurality of containers storing the container data, and based on attributes of each container; generating a priority score for each container based on the classifying; tagging each container with a priority tag based on a priority score generated for each container based on the classifying; monitoring an assignment of backup tasks to proxies for the backing up step; determining if the assignment distributes backup loads within a defined performance tolerance, and adjusting the assignment if not; and backing up the container data in an order determined by the classifying and the assignment or adjusted assignment.

Assignees

• Dell Products Lp

Inventors

• Av Mahesh Reddy
• Kumar Avinash
• O'Callaghan Terry

Classifications

• G06F11/1453

  using de-duplication of the data · CPC title

• G06F11/1464 Primary

  for networked environments · CPC title

• G06F11/1451 Primary

  by selection of backup contents · CPC title