Technology for logging levels and transaction log files

A computer system for dynamic logging, the system comprising:

• a processor;
• and
• a computer readable storage medium connected to the processor, wherein the computer readable storage medium has stored thereon a program for controlling the processor, and wherein the processor is operative with the program to execute the program for:
• generating parsed event data by at least one natural language (NL) processor responsive to event data of a log for transactions of a target application;
• classifying, by a first classifier applying a first machine learning model to the parsed event data, context states of the respective transactions of the target application;
• classifying, by a second classifier applying a second machine learning model to the parsed event data and the context states, trouble prone states of the respective transactions, wherein the trouble prone states are for respective hierarchical levels;
• determining, by a logic module, for a current one of the trouble prone states for a current transaction, that the current trouble prone state is a higher trouble prone level than for a transaction immediately preceding the current transaction;
• and
• sending, by the logic module to the target application, an increased log detail selection responsive to the determining that the current trouble prone state is the higher trouble prone level, so that a greater amount of log detail is logged for at least a next transaction after the current transaction.

A computer program product for dynamic logging, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processor to cause the processor to:

• generate parsed event data by at least one natural language (NL) processor responsive to event data of a log for transactions of a target application;
• classify, by a first classifier applying a first machine learning model to the parsed event data, context states of the respective transactions of the target application;
• classify, by a second classifier applying a second machine learning model to the parsed event data and the context states, trouble prone states of the respective transactions, wherein the trouble prone states are for respective hierarchical levels;
• determine, by a logic module, for a current one of the trouble prone states for a current transaction, that the current trouble prone state is a higher trouble prone level than for a transaction immediately preceding the current transaction;
• and
• send, by the logic module to the target application, an increased log detail selection responsive to the determining that the current trouble prone state is the higher trouble prone level, so that a greater amount of log detail is logged for at least a next transaction after the current transaction.