Clustering of structured log data by key schema

1. A computer-implemented method comprising:

receiving a request to analyze a plurality of logs containing tabular data having values for a plurality of fields, the request comprising a test predicate indicating one or more conditions for the values of the plurality of fields;

creating a test set comprising logs from the plurality of logs that meet the one or more conditions;

creating a control set comprising logs from the plurality of logs that do not meet the one or more conditions;

determining a set of key-value pairs comprising key-value pairs that occur in the test set with a frequency exceeding a first predetermined frequency and that occur in the control set with a frequency lower than a second predetermined frequency; and

causing presentation on a user interface (UI) of an explanation parameter comprising the set of key-value pairs.

2. The method as recited in claim 1 , wherein the causing presentation of the explanation parameter comprises:

providing, in the UI, a percentage of relevance based on how often each determined key-value pair appears in the test set and in the control set.

3. The method as recited in claim 1 , wherein the causing presentation of the explanation parameter comprises:

providing, in the UI, a count of how many times the determined key-value pair appears in the test set and in the control set.

4. The method as recited in claim 1 , further comprising:

providing options in the UI to select operators for defining the test predicate.

5. The method as recited in claim 1 , wherein the tabular data comprises rows for the plurality of logs and columns for parsed values extracted from structured data in the plurality of logs.

6. The method as recited in claim 1 , wherein the causing presentation of the explanation parameter comprises:

causing presentation on the UI of a list of explanations in tabular format with one row for each unique explanation parameter associated with a unique combination of values for the key-value pairs.

7. The method as recited in claim 1 , wherein the causing presentation of the explanation parameter comprises:

providing a relevance score for the explanation parameter, the relevance score calculated based on a ratio of appearance of the key-value pair between the test set and the control set.

8. The method as recited in claim 1 , further comprising:

providing an Application Programming Interface (API) for receiving the request, the API comprising configurable parameters for defining the plurality of logs, the test predicate, and a request to calculate the explanation parameter.

9. The method as recited in claim 1 , wherein the test set comprises logs indicating an abnormal condition and the control set comprises logs indicating baseline behavior.

10. A system comprising:

a memory comprising instructions; and

one or more computer processors, wherein the instructions, when executed by the one or more computer processors, cause the system to perform operations comprising:

receiving a request to analyze a plurality of logs containing tabular data having values for a plurality of fields, the request comprising a test predicate indicating one or more conditions for the values of the plurality of fields;

creating a test set comprising logs from the plurality of logs that meet the one or more conditions;

creating a control set comprising logs from the plurality of logs that do not meet the one or more conditions;

determining a set of key-value pairs comprising key-value pairs that occur in the test set with a frequency exceeding a first predetermined frequency and that occur in the control set with a frequency lower than a second predetermined frequency; and

causing presentation on a user interface (UI) of an explanation parameter comprising the set of key-value pairs.

11. The system as recited in claim 10 , wherein the causing presentation of the explanation parameter comprises:

providing, in the UI, a percentage of relevance based on how often each determined key-value pair appears in the test set and in the control set.

12. The system as recited in claim 10 , wherein the causing presentation of the explanation parameter comprises:

providing, in the UI, a count of how many times the determined key-value pair appears in the test set and in the control set.

13. The system as recited in claim 10 , wherein the instructions further cause the one or more computer processors to perform operations comprising:

providing options in the UI to select operators for defining the test predicate.

14. The system as recited in claim 10 , wherein the tabular data comprises rows for the plurality of logs and columns for parsed values extracted from structured data in the plurality of logs.

15. The system as recited in claim 10 , wherein causing presentation of the explanation parameter comprises:

causing presentation on the UI of a list of explanations in tabular format with one row for each unique explanation parameter associated with a unique combination of values for the key-value pairs.

16. A non-transitory machine-readable storage medium including instructions that, when executed by a machine, cause the machine to perform operations comprising:

receiving a request to analyze a plurality of logs containing tabular data having values for a plurality of fields, the request comprising a test predicate indicating one or more conditions for the values of the plurality of fields;

creating a test set comprising logs from the plurality of logs that meet the one or more conditions;

creating a control set comprising logs from the plurality of logs that do not meet the one or more conditions;

determining a set of key-value pairs comprising key-value pairs that occur in the test set with a frequency exceeding a first predetermined frequency and that occur in the control set with a frequency lower than a second predetermined frequency; and

causing presentation on a user interface (UI) of an explanation parameter comprising the set of key-value pairs.

17. A non-transitory machine-readable storage medium as recited in claim 16 , wherein the causing presentation of the explanation parameter comprises:

providing, in the UI, a percentage of relevance based on how often each determined key-value pair appears in the test set and in the control set.

18. A non-transitory machine-readable storage medium as recited in claim 16 , wherein the causing presentation of the explanation parameter comprises:

providing, in the UI, a count of how many times the determined key-value pair appears in the test set and in the control set.

19. A non-transitory machine-readable storage medium as recited in claim 16 , wherein the machine further performs operations comprising: providing options in the UI to select operators for defining the test predicate.

20. A non-transitory machine-readable storage medium as recited in claim 16 , wherein the tabular data comprises rows for the plurality of logs and columns for parsed values extracted from structured data in the plurality of logs.