Method and system for enabling agentless backup and restore operations on a container orchestration platform
A method and system for enabling agentless backup and restore operations on a container orchestration platform. Specifically, the method and system disclosed herein may entail facilitating the implementation of backup (or restore) workloads in a containerized environment without the co-execution of secondary storage agents therein, which traditionally manage these backup (or restore) operations.
1. A method for implementing agentless workloads, comprising:
receiving, by a master node of a cluster, a volume creation request from an administrator client, wherein the volume creation request comprises volume configuration information (VCI);
in response to the volume creation request:
generating a backup persistent volume in a persistent volume pool based on a portion of the VCI;
invoking a backup storage system (BSS) plug-in to connect the backup persistent volume to the BSS using a second portion of the VCI,
wherein the BSS is operatively connected to the cluster;
after invoking the BSS plug-in, receiving, by the master slave node, an agentless workload, sought to be implemented, alongside job configuration information;
selecting, by the master node, a slave node in the cluster to implement the agentless workload;
sending the agentless workload to the slave node;
receiving, by the slave node, the agentless workload alongside job configuration information (JCI);
extracting, from the JCI, a container image identifier (ID);
retrieving a container image associated with the container image ID; and
instantiating, to implement the agentless workload, a job pod set comprising a container created from the container image,
wherein the container executes a set of tasks to complete the agentless workload.
2. The method of claim 1 , wherein the agentless workload is a backup operation entailing a stateful application.
3. The method of claim 2 , wherein executing the set of tasks, comprises:
reading stateful application data, pertinent to the stateful application, from a cluster storage system (CSS);
replicating the stateful application data, to obtain a stateful application data copy;
pre-processing the stateful application data copy, to obtain a processed stateful application data copy; and
writing the processed stateful application data copy to the BSS.
4. The method of claim 3 , wherein reading from the CSS comprises using a cluster persistent volume mount mapped to a cluster persistent volume, wherein the cluster persistent volume is backed by the CSS.
5. The method of claim 3 , wherein writing to the BSS comprises using a backup persistent volume mount mapped to the backup persistent volume.
6. The method of claim 1 , wherein the agentless workload is a restore operation entailing a stateful application.
7. The method of claim 6 , wherein executing the set of tasks, comprises:
reading a processed stateful application data copy, pertinent to a recovery of the stateful application, from the BSS;
post-processing the processed stateful application data copy, to obtain a stateful application data copy; and
writing the stateful application data copy to a cluster storage system (CSS).
8. A system, comprising:
a slave node comprising a first computer processor;
a master node comprising a second computer processor;
a master node interface executing on the second computer processor, and programmed to:
receive a volume creation request from an administrator client, wherein the volume creation request comprises volume configuration information (VCI);
in response to the volume creation request:
generate a backup persistent volume in a persistent volume pool based on a portion of the VCI;
invoke a backup storage system (BSS) plug-in, to connect the backup persistent volume to the BSS using a second portion of the VCI,
wherein the BSS is operatively connected to the cluster;
after invoking the BSS plug-in, receive an agentless workload, sought to be implemented, alongside job configuration information;
selecting the slave node to implement the agentless workload;
sending the agentless workload to the slave node;
a container runtime executing on the first computer processor, and programmed to:
receive the agentless workload alongside job configuration information (JCI);
extract, from the JCI, a container image identifier (ID);
retrieve a container image associated with the container image ID; and
instantiate, to implement the agentless workload, a job pod set comprising a container created from the container image,
wherein the container is programmed to execute a set of tasks to complete the agentless workload.
9. The system of claim 8 , further comprising:
a container registry operatively connected to the slave node,
wherein the container image is retrieved from the container registry.
10. The system of claim 8 , further comprising:
a slave node manager also executing on the first computer processor, and operatively connected to the container runtime,
wherein the agentless workload and JCI is delegated from the slave node manager.
11. The system of claim 10 ,
wherein the agentless workload and JCI is deployed to the slave node from the master node interface.
12. The system of claim 11 , further comprising:
a plurality of slave nodes comprising the slave node, and operatively connected to the master node; and
a workload scheduler also executing on the second computer processor, and operatively connected to the master node interface,
wherein deployment of the agentless workload and JCI to the slave node, from the plurality of slave nodes, is determined by the workload scheduler using at least a set of job resource requirements (JRR) associated with the agentless workload.
13. The system of claim 12 , further comprising:
a cluster comprising the master node, the plurality of slave nodes, and a persistent volume pool,
wherein the persistent volume pool comprises a cluster persistent volume operatively connected to at least the slave node,
wherein the slave node comprises a cluster persistent volume mount through which the container accesses the cluster persistent volume to execute the set of tasks.
14. The system of claim 13 , further comprising:
a cluster storage system (CSS) operatively connected to the cluster,
wherein the cluster further comprises a CSS plug-in that maps the cluster persistent volume to the CSS.
15. The system of claim 13 , wherein the persistent volume pool further comprises the backup persistent volume operatively connected to at least the slave node, wherein the slave node further comprises a backup persistent volume mount through which the container accesses the backup persistent volume to further execute the set of tasks.
16. The system of claim 15 , further comprising:
a backup storage system (BSS) operatively connected to the cluster,
wherein the cluster further comprises the BSS plug-in.