Scalable multi-framework multi-tenant lifecycle management of deep learning applications

1. A computer-implemented lifecycle management method implemented via a processor and a memory, the memory storing instructions for the processor, the method comprising:

establishing, via the processor, a public-key infrastructure (PKI) for end-to-end encryption of control plane and data plane communications by providing encryption via public key/private key pairs between arbitrary components for application execution where an interaction pattern is specified for isolated and secure execution in a multi-tenant environment; and

outlining and establishing a network policy on-demand restricting a communication in a cluster executing a deep learning job by restricting a communication from each job component only to another job component of the same job,

wherein a cluster-level certificate authority digitally signs and distributes a public key of each component, and

wherein the network policy is dynamically added by using software defined networking (SDN) infrastructure to the cluster executing the deep learning job.

2. The computer-implemented lifecycle management method of claim 1 , wherein the multi-tenant environment includes an environment in which each tenant of the multi-tenant environment is separate from each other.

3. The computer-implemented lifecycle management method of claim 1 , further comprising automatically generating and maintaining an audit trail of the PKI for compliance.

4. The computer-implemented lifecycle management method of claim 1 , further comprising automatically selecting a parameter and a configuration of the PKI based on a desired regulatory compliance.

5. The computer-implemented lifecycle management method of claim 3 , further comprising automatically selecting a parameter and a configuration of the PKI based on a desired regulatory compliance.

6. The computer-implemented lifecycle management method of claim 1 , further comprising performing a secure delete of main memory and storage based on the encryption.

7. The computer-implemented lifecycle management method of claim 5 , further comprising performing a secure delete of main memory and storage based on the encryption.

8. The computer-implemented lifecycle management method of claim 1 , embodied in a cloud-computing environment.

9. A non-transitory computer program product for lifecycle management, the computer program product comprising a computer-readable storage medium having program instructions embodied therewith, the program instructions executable by a computer to cause the computer to perform:

establishing, via the computer, a public-key infrastructure (PKI) for end-to-end encryption of control plane and data plane communications by providing encryption via public key/private key pairs between arbitrary components for application execution where an interaction pattern is specified for isolated and secure execution in a multi-tenant environment; and

outlining and establishing a network policy on-demand restricting a communication in a cluster executing a deep learning job by restricting a communication from each job component only to another job component of the same job,

wherein a cluster-level certificate authority digitally signs and distributes a public key of each component, and

wherein the network policy is dynamically added by using software defined networking (SDN) infrastructure to the cluster executing the deep learning job.

10. The non-transitory computer program product for lifecycle management of claim 9 , wherein the multi-tenant environment includes an environment in which each tenant of the multi-tenant environment is separate from each other.

11. The non-transitory computer program product for lifecycle management of claim 9 , further comprising automatically generating and maintaining an audit trail of the PKI for compliance.

12. The non-transitory computer program product for lifecycle management of claim 9 , further comprising automatically selecting a parameter and a configuration of the PKI based on a desired regulatory compliance.

13. The non-transitory computer program product for lifecycle management of claim 11 , further comprising automatically selecting a parameter and a configuration of the PKI based on a desired regulatory compliance.

14. The non-transitory computer program product for lifecycle management of claim 9 , further comprising performing a memory-secure-delete based on the encryption.

15. A lifecycle management system, the system comprising:

a processor; and

a memory, the memory storing instructions to cause the processor to perform:

establishing a public-key infrastructure (PKI) for end-to-end encryption of control plane and data plane communications by providing encryption via public key/private key pairs between arbitrary components for application execution where an interaction pattern is specified for isolated and secure execution in a multi-tenant environment; and

outlining and establishing a network policy on-demand restricting a communication in a cluster executing a deep learning job by restricting a communication from each job component only to another job component of the same job,

wherein a cluster-level certificate authority digitally signs and distributes a public key of each component, and

wherein the network policy is dynamically added by using software defined networking (SDN) infrastructure to the cluster executing the deep learning job.

16. The computer-implemented lifecycle management method of claim 1 , wherein the distribution of the public key is from each component to every other component of the arbitrary components.

17. The computer-implemented lifecycle management method of claim 1 , wherein the network policy restricts access in the cluster between each tenant of the multi-tenant environment.

18. The computer-implemented lifecycle management method of claim 2 , wherein the network policy restricts access in the cluster between each tenant of the multi-tenant environment.