The sample dataset Apache contains the files access.log and error.log that contains
the logfile of the accesses to a web server and the errors.
The access.log is in Common Log
Format.
The entries in error.log usually have a corresponding entry in access.log
- Read the file access.log
- Count the number of accesses (number of lines) made by an IP number (an IP consist of four numbers separated by a dot)
- Count the number of successful accesses (status 200) made by an IP number
- Count the number of accesses for each directory served (the url consists of a directory and a file name)
- For each origin, count the number of successful accesses
- For each origin, count the number of unsuccessful accesses, split according to the
status code
- From the results of the previous point, add a column with the error class (the first
digit of the status code)
- Cluster the accesses in 5-minutes time slices (e.g. from 14:00 to 14:05, from 14:05 to
14:10, etc). Count the number of accesses for each time slice
- For each
[info] entry of error.log, find the next entry of access.log. For
example, when considering the entry at Sun Mar 7 18:00:09 2004, we want to find the
entry at [07/Mar/2004:18:02:10 -0800]
- Count the number of times that the two accesses of the previous point have the same origin.
- Count the number of accesses between each pair of
[info] or [error] entries of error.log