Adaptive system for smart boot sequence formation of VMs for disaster recovery
One example method includes receiving input concerning a boot order sequence, where the input includes VM metadata, entering a training phase which includes generating a boot sequence rule based on the input, using the boot sequence rule to generate a proposed boot sequence, performing the proposed boot sequence, and gathering information concerning performance of the proposed boot sequence. The gathered information can be used as a basis to generate a modified boot sequence.
1. A method, comprising the following operations:
providing training data to a boot sequence algorithm, and the training data comprises VM (virtual machine) metadata;
extracting features from the training data, and providing the features to the boot sequence algorithm;
providing boot sequence information to the boot sequence algorithm;
establishing, by the boot sequence algorithm, an initial set of rules for determining a boot sequence, and the initial set of rules is established based on the training data, features, and boot sequence information;
generating, by the boot sequence algorithm, a proposed boot sequence, and the proposed boot sequence is generated based at least in part on the initial set of rules;
associating the proposed boot sequence with results expected to be obtained by implementation of the proposed boot sequence;
testing the proposed boot sequence; and
collecting real world telemetry data of actual performance obtained using the proposed boot sequence.
2. The method as recited in claim 1 , wherein the VM metadata includes any one or more of: types of applications installed on the VM; VM uptime statistics; VM network and storage configuration information; VM tags; VM naming conventions; and, VM labels.
3. The method as recited in claim 1 , wherein the training data is used to perform an initial training process for the boot sequence algorithm.
4. The method as recited in claim 1 , wherein the boot sequence algorithm is part of a machine learning system that is operable to automatically generate and optimize a boot sequence based at least in part on the proposed boot sequence.
5. The method as recited in claim 1 , wherein the boot sequence information was generated without human input.
6. The method as recited in claim 1 , wherein the boot sequence information was previously used by the boot sequence algorithm.
7. The method as recited in claim 1 , wherein the boot sequence information comprises a default boot sequence that constitutes a best initial guess as to what an optimized boot sequence should be.
8. The method as recited in claim 1 , wherein the real world telemetry comprises real world results of running the proposed boot sequence, and the operations further comprise comparing the real world results with the expected results, and modifying the proposed boot sequence if necessary.
9. The method as recited in claim 1 , wherein testing the proposed boot order comprises running an orchestrated DR (disaster recovery) test.
10. The method as recited in claim 1 , wherein the operations further comprise performing a disaster recovery operation using a boot sequence based at least in part on the proposed boot sequence.
11. A non-transitory computer readable storage medium having stored therein instructions which are executable by one or more hardware processors to perform operations comprising:
providing training data to a boot sequence algorithm, and the training data comprises VM (virtual machine) metadata;
extracting features from the training data, and providing the features to the boot sequence algorithm;
providing boot sequence information to the boot sequence algorithm;
establishing, by the boot sequence algorithm, an initial set of rules for determining a boot sequence, and the initial set of rules is established based on the training data, features, and boot sequence information;
generating, by the boot sequence algorithm, a proposed boot sequence, and the proposed boot sequence is generated based at least in part on the initial set of rules;
associating the proposed boot sequence with results expected to be obtained by implementation of the proposed boot sequence;
testing the proposed boot sequence; and
collecting real world telemetry data of actual performance obtained using the proposed boot sequence.
12. The non-transitory computer readable storage medium as recited in claim 11 , wherein the VM metadata includes any one or more of:
types of applications installed on the VM; VM uptime statistics; VM network and storage configuration information; VM tags; VM naming conventions; and, VM labels.
13. The non-transitory computer readable storage medium as recited in claim 11 , wherein the training data is used to perform an initial training process for the boot sequence algorithm.
14. The non-transitory computer readable storage medium as recited in claim 11 , wherein the boot sequence algorithm is part of a machine learning system that is operable to automatically generate and optimize a boot sequence that is based at least in part on the proposed boot sequence.
15. The non-transitory computer readable storage medium as recited in claim 11 , wherein the boot sequence information was generated without human input.
16. The non-transitory computer readable storage medium as recited in claim 11 , wherein the boot sequence information was previously used by the boot sequence algorithm.
17. The non-transitory computer readable storage medium as recited in claim 11 , wherein the boot sequence information comprises a default boot sequence that constitutes a best initial guess as to what an optimized boot sequence should be.
18. The non-transitory computer readable storage medium as recited in claim 11 , wherein the real world telemetry comprises real world results of running the proposed boot sequence, and the operations further comprise comparing the real world results with the expected results, and modifying the proposed boot sequence if necessary.
19. The non-transitory computer readable storage medium as recited in claim 11 , wherein testing the proposed boot order comprises running an orchestrated DR (disaster recovery) test.
20. The non-transitory computer readable storage medium as recited in claim 11 wherein the operations further comprise performing a disaster recovery operation using a boot sequence that is based at least in part on the proposed boot sequence.