What are the meanings of file names in the Spark dataset?

[LITONG99]

Hi, thank you for the valuable resources.

According to the paper, Spark logs are aggregated at the machine level and are from 32 machines. However, there are 194 directories in the data folder, with a sequence of numbers as the identifier. Under each directory, there are an arbitrary number of files with a sequence of numbers as the identifier.

Spark/application_1485248649253_0170/container_1485248649253_0170_01_000002.log

Could you please provide me with more details concerning:

1. What do the numbers in the file name mean? For example, "1485248649253", "170", "01", "000002".
2. How can I separate the logs from the 32 machines?
3. Which version of Spark do you use to generate the logs?

Thank you! I appreciate any response :)