Kyndryl, Inc.

Ingo Averdunk

Harish Bharti

Rajesh Kumar Saxena

Vishal Anand

Shankar S Kalyana

Rajib Das

A method, computer program product, and system include a processor(s) that obtains a request for a transaction to be processed by an application in a computing system. The processor(s) applies, to the request for the transaction, an outlier detection model, to determine whether the transaction comprises attributes matching transaction tuples of one or more historical transactions identified as triggering issues in the computing system when the application processed the historical transactions. The processor(s) classifies the transaction as being an outlier transaction or as being a standard transaction, based on applying the outlier detection model. Based on determining that the transaction is an outlier transaction, concurrently with the application processing the transaction, the processor(s) turn on the debug trace to debug trace the application processing of the transaction.

Dynamic debug tracing with machine learning

A system, computer program product, and method are presented for automatically executing chaos experiments on computing resources, applications, and services through automatically establishing and meeting core requirements for each chaos experiment. The method includes receiving a trigger signal configured to establish one or more condition s to execute one or more chaos experiments on at least a portion of one or more landing zones. The one or more chaos experiments are configured to operationally stress one or more of one or more infrastructure resources, one or more services, and one or more computing applications. The method also includes determining, automatically, the technical support coverage for the execution of the one or more chaos experiments. The method further includes executing, automatically, subject to the technical support coverage determination, the one or more chaos experiments.

Growing community of inventors

Markt Schwaben, Germany

Ingo Averdunk