Machine learning-driven orchestration engine for managing application programming interface failures in container-based data protection systems

A non-transitory computer readable medium comprising computer readable program code, which when executed by a computer processor enables the computer processor to perform a method for managing a data protection system, the method comprising:

• monitoring, by an interaction manager, a data protection workload executing on the data protection system,
• wherein the data protection system comprises a plurality of services, and
• wherein the data protection workload is executed by the plurality of services communicating with each other using application programming interface (API) calls;
• obtaining, based on the monitoring, a copy of the API calls;
• performing an embedding on the copy of the API calls to obtain API embeddings;
• applying the API embeddings to a failure prediction engine to obtain a failure prediction for the data protection workload;
• performing a remediation action on the API calls to remediate the data protection workload,
• wherein the remediation action comprises at least one of:
• re-initiating the data protection workload and replacing one of the plurality of services;
• updating the failure prediction engine based on the remediation action;
• obtaining a set of historical API calls between the plurality of services for previous data protection workloads;
• performing an API sequence analysis on the set of historical API calls to obtain second API embeddings of a sequencing of the set of historical API calls;
• performing a payload similarity analysis on payloads of each of the set of historical API calls to obtain a similarity model of the set of historical API calls;
• performing a service log analysis on services logs associated with the previous data protection workloads, wherein the service logs each indicate a result of each of the previous data protection workloads;
• and
• performing a training on the failure prediction engine using the similarity model, the second API embeddings, and the service log analysis.

A system comprising:

• an interaction manager, operating on a processor;
• and
• memory comprising instructions, which when executed by the processor, perform a method comprising:
• monitoring, by an interaction manager, a data protection workload executing on a data protection system,
• wherein the data protection system comprises a plurality of services, and
• wherein the data protection workload is executed by the plurality of services communicating with each other using application programming interface (API) calls;
• obtaining, based on the monitoring, a copy of the API calls;
• performing an embedding on the copy of the API calls to obtain API embeddings;
• applying the API embeddings to a failure prediction engine to obtain a failure prediction for the data protection workload;
• performing a remediation action on the API calls to remediate the data protection workload,
• wherein the remediation action comprises at least one of:
• re-initiating the data protection workload and replacing one of the plurality of services;
• updating the failure prediction engine based on the remediation action;
• obtaining a set of historical API calls between the plurality of services for previous data protection workloads;
• performing an API sequence analysis on the set of historical API calls to obtain second API embeddings of a sequencing of the set of historical API calls;
• performing a payload similarity analysis on payloads of each of the set of historical API calls to obtain a similarity model of the set of historical API calls;
• performing a service log analysis on services logs associated with the previous data protection workloads, wherein the service logs each indicate a result of each of the previous data protection workloads;
• and
• performing a training on the failure prediction engine using the similarity model, the second API embeddings, and the service log analysis.